My session to the DECUServe system (Eisner) terminated last evening.
Pings and traceroutes seem to indicate that it is down?  Does anyone
know if it is network or system/hardware related?  I don't have any
contact info handy for Stephen Arnold to query him.

-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

  "Well my son, life is like a beanstalk, isn't it?"

In article <00A846F0.5F9DC58F@SendSpamHere.ORG>, VAXman-  @SendSpamHere.ORG writes:
>My session to the DECUServe system (Eisner) terminated last evening.
>Pings and traceroutes seem to indicate that it is down?  Does anyone
>know if it is network or system/hardware related?  I don't have any
>contact info handy for Stephen Arnold to query him.

It would appear that the machine itself of down.  Stephen Arnold's web site
(*.*.*.30) is up and it is on the same subnet as Eisner (*.*.*.34).  I hope
will be back up soon.


-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

  "Well my son, life is like a beanstalk, isn't it?"

Any update on this?  I have all of my e-mail forwarded to this machine
(probably not so smart thing to do at the first place, but I just
liked how PMAS was handling my spam), cannot access it for 2 days in a
row now.  I was thinking the machine will be back today, but now I'm
afraid it will be offline throughout all of holidays...

In article <b0f5b4cc-0f86-4b56-9a82-b09606a8dd1f@k1g2000prb.googlegroups.com>, sagrailo@gmail.com writes:
>Any update on this?  I have all of my e-mail forwarded to this machine
>(probably not so smart thing to do at the first place, but I just
>liked how PMAS was handling my spam), cannot access it for 2 days in a
>row now.  I was thinking the machine will be back today, but now I'm
>afraid it will be offline throughout all of holidays...

I sent an email to Stephen Arnold but I have received no correspodence back
from him.  I do not have any other contact info for him.  If I do hear from
him, I will report here.


-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

  "Well my son, life is like a beanstalk, isn't it?"

I also tried sending email, and have not received a response.  Perhaps he is 
on vacation?

His regular phone number may be  (608) 278 7700

<VAXman- @SendSpamHere.ORG> wrote in message 
news:00A847C1.2583DEE3@SendSpamHere.ORG...
> In article 
> <b0f5b4cc-0f86-4b56-9a82-b09606a8dd1f@k1g2000prb.googlegroups.com>, 
> sagrailo@gmail.com writes:
>>Any update on this?  I have all of my e-mail forwarded to this machine
>>(probably not so smart thing to do at the first place, but I just
>>liked how PMAS was handling my spam), cannot access it for 2 days in a
>>row now.  I was thinking the machine will be back today, but now I'm
>>afraid it will be offline throughout all of holidays...
>
> I sent an email to Stephen Arnold but I have received no correspodence 
> back
> from him.  I do not have any other contact info for him.  If I do hear 
> from
> him, I will report here.
>
>
> -- 
> VAXman- A Bored Certified VMS Kernel Mode Hacker 
> VAXman(at)TMESIS(dot)ORG
>
>  "Well my son, life is like a beanstalk, isn't it?" 

DeCoy wrote:
> I also tried sending email, and have not received a response.  Perhaps he is 
> on vacation?
> 
> His regular phone number may be [...]

Hi, Dale! Don't think I have ever seen you post here, at least not using
this particular nym.

How would I know how to cantact anyone?  Evenything I need to survive
is on eisner!

Not really true, but sometimes it seems like it. If you use a site
that has been stable for almost 20 years, you start thnking it will
be there forever.

Hopefully, this is just a temporary thing.  Steve can't be home _all_
of the time just to keep a server up for us.

[With any luck this will show up OK - I'm not used to posting via
Thunderbird or via this server].

-- 
George Cornelius                cornelius at mayo dot edu
                                cornelius at encompasserve dot org

"G Cornelius" <cornelius@eisner.decus.invalid> wrote in message 
news:49509de3$0$89393$815e3792@news.qwest.net...
> DeCoy wrote:
>> I also tried sending email, and have not received a response.  Perhaps he 
>> is
>> on vacation?
>>
>> His regular phone number may be [...]
>
> Hi, Dale! Don't think I have ever seen you post here, at least not using
> this particular nym.
>

Hi, George - and others.  I usually don't have any big reason to post to the 
group.

I just spoke to Steve Arnold - who was in an airport, heading home.  When he 
gets there, he'll find out what's up.

We can hope that it's something simple. 

In article <u_84l.15339$ZP4.4398@nlpi067.nbdc.sbc.com>, "DeCoy" <daleecoy@obfuscation.spinn.net> writes:
>
>"G Cornelius" <cornelius@eisner.decus.invalid> wrote in message 
>news:49509de3$0$89393$815e3792@news.qwest.net...
>> DeCoy wrote:
>>> I also tried sending email, and have not received a response.  Perhaps he 
>>> is
>>> on vacation?
>>>
>>> His regular phone number may be [...]
>>
>> Hi, Dale! Don't think I have ever seen you post here, at least not using
>> this particular nym.
>>
>
>Hi, George - and others.  I usually don't have any big reason to post to the 
>group.
>
>I just spoke to Steve Arnold - who was in an airport, heading home.  When he 
>gets there, he'll find out what's up.
>
>We can hope that it's something simple. 

I think all EISNER users hope it is something simple.


-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

  "Well my son, life is like a beanstalk, isn't it?"

On Dec 21, 6:30=A0pm, VAXman-  @SendSpamHere.ORG wrote:
> My session to the DECUServe system (Eisner) terminated last evening.
> Pings and traceroutes seem to indicate that it is down? =A0Does anyone
> know if it is network or system/hardware related? =A0I don't have any
> contact info handy for Stephen Arnold to query him.

I've been paying a modest contribution for Eisner for the last few
years,
but admittedly did not help yet this year. Didn't expect the effect to
be immediate!

Eisner being down for an extended period perhaps short be interpreted
as a strong reminder to all who use it to chip in a little more, a
little more often, ( personal circumstances permitting).

Cheers,
Hein.
(Writing from Eindhoven, the Netherlands, today)

In article <00A847C1.2583DEE3@SendSpamHere.ORG>, VAXman-  @SendSpamHere.ORG writes:
>In article <b0f5b4cc-0f86-4b56-9a82-b09606a8dd1f@k1g2000prb.googlegroups.com>, sagrailo@gmail.com writes:
>>Any update on this?  I have all of my e-mail forwarded to this machine
>>(probably not so smart thing to do at the first place, but I just
>>liked how PMAS was handling my spam), cannot access it for 2 days in a
>>row now.  I was thinking the machine will be back today, but now I'm
>>afraid it will be offline throughout all of holidays...
>
>I sent an email to Stephen Arnold but I have received no correspodence back
>from him.  I do not have any other contact info for him.  If I do hear from
>him, I will report here.

Eisner will be down 10 days as of tomorrow.  I've still not heard anything 
about what has happened with this system.  I, and many others, would sure
like to see it back on-line soon.

-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

  "Well my son, life is like a beanstalk, isn't it?"

<VAXman- @SendSpamHere.ORG> wrote in message 
news:00A84D7E.FD3E6654@SendSpamHere.ORG...
> In article <00A847C1.2583DEE3@SendSpamHere.ORG>, VAXman- 
> @SendSpamHere.ORG writes:
>>In article 
>><b0f5b4cc-0f86-4b56-9a82-b09606a8dd1f@k1g2000prb.googlegroups.com>, 
>>sagrailo@gmail.com writes:
>>>Any update on this?  I have all of my e-mail forwarded to this machine
>>>(probably not so smart thing to do at the first place, but I just
>>>liked how PMAS was handling my spam), cannot access it for 2 days in a
>>>row now.  I was thinking the machine will be back today, but now I'm
>>>afraid it will be offline throughout all of holidays...
>>
>>I sent an email to Stephen Arnold but I have received no correspodence 
>>back
>>from him.  I do not have any other contact info for him.  If I do hear 
>>from
>>him, I will report here.
>
> Eisner will be down 10 days as of tomorrow.  I've still not heard anything
> about what has happened with this system.  I, and many others, would sure
> like to see it back on-line soon.
>
> -- 
> VAXman- A Bored Certified VMS Kernel Mode Hacker 
> VAXman(at)TMESIS(dot)ORG
>
>  "Well my son, life is like a beanstalk, isn't it?"

I have not heard anything, either, since I spoke to Steve (when he was in an 
airport) last week.

As you know, Eisner is housed in Steve's residence.  There's a "spare" DS20 
for parts, and I believe that backups were good (and automatic).

I'm presuming that Steve will get the system fixed and online as quickly as 
he can.  Or that he will ask for help from myself or others, if he needs it.

If my presumptions are not correct, then there are not many practical 
alternatives.

Steve can be called at (608) 278-7700 if you would like to determine the 
current situation (and repeat it here), or if someone is close enough to 
Madison, Wisconsin to offer hands-on assistance.

Needless to say, I share your desires and hopes to see it back online soon. 

DeCoy wrote:

> As you know, Eisner is housed in Steve's residence. 


Wasn't it originally in DECUS/Digital premises ? How long ago did it
become a residential based system ? What sort of iternet connection does
it have ? (DSL, Cable, or some business technology with higher speed ?)

In article <0049bf98$0$13622$c3e8da3@news.astraweb.com>, JF Mezei <jfmezei.spamnot@vaxination.ca> writes:
>DeCoy wrote:
>
>> As you know, Eisner is housed in Steve's residence. 
>
>
>Wasn't it originally in DECUS/Digital premises ? How long ago did it
>become a residential based system ? What sort of iternet connection does
>it have ? (DSL, Cable, or some business technology with higher speed ?)

I believe that prior to the move to Stephen's residence that it was
housed in Terry Kennedy's residence in Jersey City, NJ.  I managed
to get some of the bits and pieces of the original Eisner.

-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

  "Well my son, life is like a beanstalk, isn't it?"

"JF Mezei" <jfmezei.spamnot@vaxination.ca> wrote in message 
news:0049bf98$0$13622$c3e8da3@news.astraweb.com...
> DeCoy wrote:
>
>> As you know, Eisner is housed in Steve's residence.
>
>
> Wasn't it originally in DECUS/Digital premises ?

Yes

>How long ago did it
> become a residential based system ?

Approximately, when DECUS became Encompass.

>What sort of iternet connection does
> it have ? (DSL, Cable, or some business technology with higher speed ?)

Steve's LAN has a T1 connection. 

I just got a couple emails, indicating that I've been added to an
EISNER group at arnold dot com. I never requested this, but I'm
guessing that Steve has been away for the holidays, and is finally
getting back to where he can deal with whatever happened.

Like everything else, he's a volunteer, not a 24 x 365 employee.
Thanks for keeping EISNER available all these years, and I hope it's
back online soon.

In article <039e00ba-6aee-4b5b-b22e-aff619e1bf09@s1g2000prg.googlegroups.com>, Bob <bobkaplow@yahoo.com> writes:
>I just got a couple emails, indicating that I've been added to an
>EISNER group at arnold dot com. I never requested this, but I'm
>guessing that Steve has been away for the holidays, and is finally
>getting back to where he can deal with whatever happened.

I've gotten nothing and Eisner is still unreachable.


>Like everything else, he's a volunteer, not a 24 x 365 employee.
>Thanks for keeping EISNER available all these years, and I hope it's
>back online soon.

I think that those of us who were daily denizens on Eisner would just 
like to know what happened and, if necessary, what we can do to bring
it back on-line.

-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

  "Well my son, life is like a beanstalk, isn't it?"

VAXman- @SendSpamHere.ORG wrote:

> I've gotten nothing and Eisner is still unreachable.

Your left handed SMTP server often bounces messages left and right :-)
:-) Quite possible that a message sent to you got bounced.

> I think that those of us who were daily denizens on Eisner would just 
> like to know what happened and, if necessary, what we can do to bring
> it back on-line.


You need to drink a half dozen Guiness, kneel down, and perform
religious incantations to bring it back.

In article <00483066$0$11622$c3e8da3@news.astraweb.com>, JF Mezei <jfmezei.spamnot@vaxination.ca> writes:
>VAXman- @SendSpamHere.ORG wrote:
>
>> I've gotten nothing and Eisner is still unreachable.
>
>Your left handed SMTP server often bounces messages left and right :-)
>:-) Quite possible that a message sent to you got bounced.

Nothing in the logs to indicate that.



>> I think that those of us who were daily denizens on Eisner would just 
>> like to know what happened and, if necessary, what we can do to bring
>> it back on-line.
>
>
>You need to drink a half dozen Guiness, kneel down, and perform
>religious incantations to bring it back.

The dozen Guinness -- doable.
Kneel down on my knees?  Not likely...
Religious Incantations?  Perhaps I could muster up an atheistic rant.
-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

  "Well my son, life is like a beanstalk, isn't it?"

DeCoy wrote:
> 
> I have not heard anything, either, since I spoke to Steve (when he was in an 
> airport) last week.
> 
> As you know, Eisner is housed in Steve's residence.  There's a "spare" DS20 
> for parts, and I believe that backups were good (and automatic).
> 
> I'm presuming that Steve will get the system fixed and online as quickly as 
> he can.  Or that he will ask for help from myself or others, if he needs it.
> 
> If my presumptions are not correct, then there are not many practical 
> alternatives.
> 
> Steve can be called at (608) 278-7700 if you would like to determine the 
> current situation (and repeat it here), or if someone is close enough to 
> Madison, Wisconsin to offer hands-on assistance.

I'm 3-4 hours away, and am willing to help. Might even have a spare DS20
available. Unfortuneately, starting Tuesday I'm on a short tether (pager)
for about a week.

-- 
George Cornelius                       cornelius at mayo dot edu

In article <4960b9bf$0$48218$815e3792@news.qwest.net>, G Cornelius <cornelius@eisner.decus.invalid> writes:
>DeCoy wrote:
>> 
>> I have not heard anything, either, since I spoke to Steve (when he was in an 
>> airport) last week.
>> 
>> As you know, Eisner is housed in Steve's residence.  There's a "spare" DS20 
>> for parts, and I believe that backups were good (and automatic).
>> 
>> I'm presuming that Steve will get the system fixed and online as quickly as 
>> he can.  Or that he will ask for help from myself or others, if he needs it.
>> 
>> If my presumptions are not correct, then there are not many practical 
>> alternatives.
>> 
>> Steve can be called at (608) 278-7700 if you would like to determine the 
>> current situation (and repeat it here), or if someone is close enough to 
>> Madison, Wisconsin to offer hands-on assistance.
>
>I'm 3-4 hours away, and am willing to help. Might even have a spare DS20
>available. Unfortuneately, starting Tuesday I'm on a short tether (pager)
>for about a week.

Stephen sent an email out to a distribution list.  I read it late last
evening (actually, early this morning ~2:30).  The problem appears to
be in the RAID configuration, a drive in the RAID configuration or its
controller.

-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

  "Well my son, life is like a beanstalk, isn't it?"

VAXman- @SendSpamHere.ORG wrote:
> In article <4960b9bf$0$48218$815e3792@news.qwest.net>, G Cornelius <cornelius@eisner.decus.invalid> writes:
>> DeCoy wrote:
>>> I have not heard anything, either, since I spoke to Steve (when he was in an 
>>> airport) last week.
<snip>
> 
> Stephen sent an email out to a distribution list.  I read it late last
> evening (actually, early this morning ~2:30).  The problem appears to
> be in the RAID configuration, a drive in the RAID configuration or its
> controller.
> 

Why does the idea of the failure of a drive in a "RAID configuration" 
sound slightly bogus?  RAID 1 or RAID 5 shouldn't be bothered by the 
failure of a single drive!  Two failing drives would bring down most 
RAID configurations.  A really robust RAID 1 configuration would have 
dual controllers.

VAXman- @SendSpamHere.ORG wrote:
> 
> Stephen sent an email out to a distribution list.  I read it late last
> evening (actually, early this morning ~2:30).  The problem appears to
> be in the RAID configuration, a drive in the RAID configuration or its
> controller.

Can you forward that e-mail wb8tyw at my gmail domain address?

Apparently you are not accepting e-mail from gmail at this time.

-John
wb8tyw@gmail.company
Personal Opinion Only

John E. Malmberg wrote:
> VAXman- @SendSpamHere.ORG wrote:
>>
>> Stephen sent an email out to a distribution list.  I read it late last
>> evening (actually, early this morning ~2:30).  The problem appears to
>> be in the RAID configuration, a drive in the RAID configuration or its
>> controller.
> 
> Can you forward that e-mail wb8tyw at my gmail domain address?
> 
> Apparently you are not accepting e-mail from gmail at this time.
> 
> -John
> wb8tyw@gmail.company
> Personal Opinion Only

VAXman's spam filter needs a codeword in the Subject line in order to 
let your mail through.  I don't remember what it is.  I suppose I could 
persuade my "auxilliary brain" to cough it up but it's not mine to publish.

"G Cornelius" <cornelius@eisner.decus.invalid> wrote in message 
news:4960b9bf$0$48218$815e3792@news.qwest.net...
>
> I'm 3-4 hours away, and am willing to help. Might even have a spare DS20
> available. Unfortuneately, starting Tuesday I'm on a short tether (pager)
> for about a week.
>
> -- 
> George Cornelius                       cornelius at mayo dot edu
>

Thanks, George.  The problem appears to be storage-related, perhaps with the 
RAID array on the Mylex controller, and perhaps with either controller 
hardware or controller configuration.

Expertise in diagnosing (and perhaps fixing) Mylex controller symptoms would 
be initially useful.

DeCoy wrote:
> "G Cornelius" <cornelius@eisner.decus.invalid> wrote in message 
> news:4960b9bf$0$48218$815e3792@news.qwest.net...
>> I'm 3-4 hours away, and am willing to help. Might even have a spare DS20
>> available. Unfortuneately, starting Tuesday I'm on a short tether (pager)
>> for about a week.
>>
>> -- 
>> George Cornelius                       cornelius at mayo dot edu
>>
> 
> Thanks, George.  The problem appears to be storage-related, perhaps with the 
> RAID array on the Mylex controller, and perhaps with either controller 
> hardware or controller configuration.
> 
> Expertise in diagnosing (and perhaps fixing) Mylex controller symptoms would 
> be initially useful.

The mylex controller can not handle logical volumes greater than a 
certain size.  If I recall correctly it was either 9 or 18 GB.  It has 
been a long time.

This leads to a restriction on the size of the physical replacement drives.

I have seen the use of larger (unsupported) drives as long as the size 
of the resulting logical volumes are below the size limitation of the 
controller.

The RCU utility is on the ALPHA firmware disk, and must be run from the 
console mode.

The ALPHA flavor of the mylex controller is different than the Intel 
version at least at the firmware level.  I do not know if they can be 
interchanged with just an adjustment to the firmware.

I only used the Mylex controller on one VMS/Alpha system, and that was 
quite some time ago.

I think battery backed up cache was an option on that controller.

-John
wb8tyw@gmail.company
Personal Opinion Only

In article <Uz58l.454787$TT4.83297@attbi_s22>,
	"John E. Malmberg" <wb8tyw@qsl.network> writes:
> DeCoy wrote:
>> "G Cornelius" <cornelius@eisner.decus.invalid> wrote in message 
>> news:4960b9bf$0$48218$815e3792@news.qwest.net...
>>> I'm 3-4 hours away, and am willing to help. Might even have a spare DS20
>>> available. Unfortuneately, starting Tuesday I'm on a short tether (pager)
>>> for about a week.
>>>
>>> -- 
>>> George Cornelius                       cornelius at mayo dot edu
>>>
>> 
>> Thanks, George.  The problem appears to be storage-related, perhaps with the 
>> RAID array on the Mylex controller, and perhaps with either controller 
>> hardware or controller configuration.
>> 
>> Expertise in diagnosing (and perhaps fixing) Mylex controller symptoms would 
>> be initially useful.
> 
> The mylex controller can not handle logical volumes greater than a 
> certain size.  If I recall correctly it was either 9 or 18 GB.  It has 
> been a long time.
> 
> This leads to a restriction on the size of the physical replacement drives.

So, what kind/size of replacement would he be looking for?  I have a number
small SCSI DEC re-badged Seagates that I use in my StorageWorks Array.  We
are talking 2GB here.  Would this be what he may be looking for?  I could
probably be persuaded to donate a drive (or a couple, even) in return for
an Eisner account.  :-)

bill

-- 
Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolves
billg999@cs.scranton.edu |  and a sheep voting on what's for dinner.
University of Scranton   |
Scranton, Pennsylvania   |         #include <std.disclaimer.h>   

Bill Gunshannon wrote:
> In article <Uz58l.454787$TT4.83297@attbi_s22>,
> 	"John E. Malmberg" <wb8tyw@qsl.network> writes:
>> The mylex controller can not handle logical volumes greater than a 
>> certain size.  If I recall correctly it was either 9 or 18 GB.  It has 
>> been a long time.
>>
>> This leads to a restriction on the size of the physical replacement drives.
> 
> So, what kind/size of replacement would he be looking for?  I have a number
> small SCSI DEC re-badged Seagates that I use in my StorageWorks Array.  We
> are talking 2GB here.  Would this be what he may be looking for?  I could
> probably be persuaded to donate a drive (or a couple, even) in return for
> an Eisner account.  :-)

I am only going from memory here, but as I recall, the maximum supported 
   disk was 18 GB for a logical volume Raid 5 set of 45 Mb after formatting.

-John
wb8tyw@gmail.company
Personal Opinion Only

DeCoy wrote:

[...]
> Thanks, George.  The problem appears to be storage-related, perhaps with the 
> RAID array on the Mylex controller, and perhaps with either controller 
> hardware or controller configuration.
>
> Expertise in diagnosing (and perhaps fixing) Mylex controller symptoms would 
> be initially useful.
>
>   
I'm assuming, of course, that a recent backup of the system (or at least 
the data on the RAID array) is available.
I have a number of "bare" 18 and 36G drives that I'm willing to donate 
to the "cause", if the decision comes to abandoning the RAID array, and 
re-building storage with HBVS, instead.  The current OS would support 
"shadowing" of dissimilar-sized disks, and disks of this size (or 
larger) should do away with the "need" for the RAID array.
[...]

To answer a number of replies:

I have suggested that the easiest and most straightforward solution is to 
get the Mylex-driven RAID 5 array to work.

We have spare 9GB drives for that.  My suggestion was that Stephen, with 
whatever support and advice are available, work on that FIRST, to see if 
it's viable (given hardware, software, and peopleware).

If we can't do that, then we'll explore alternatives for replacing the whole 
disk setup.  We may have some in hand, but offers are appreciated.  We would 
certainly go to bigger drives, at that point.

I'll briefly quote Stephen here, in case somebody can spot something:

<quote>
In some of tries of the last experiment, the system console came alive
and attempted to boot VMS.  You can see the last of the output in the
screen shot at http://www.Arnold.com/Photo_123108_002.jpg.  Note the
message "SCSI drive at channel 2, target 1 dead".  The message about CPU
01 starting are not relevant.  Nothing else happens and there is no disk
activity after that message.


From this point I was able to ^P and give console commands.  Since
storage is the problem, I entered "show disk".  The output is in the
screen shot at http://www.Arnold.com/Photo_123108_003.jpg.  DRA0-5 are
drives defined on the seven-member RAID set in the built-in shelf, on
channels 1-2 of the Mylex controller.  DRA6 is on the six-member RAID
set in the external gray storage shelf.  (We also have the FTP
repository on large ATA drive connected via an ATA-SCSI adpater: DKA0.)
<end of quote>



[Footnote: somebody mentioned getting an account on Eisner - those are 
"free", of course assuming that the system runs] 

In article <Kra8l.10393$D32.329@flpi146.ffdc.sbc.com>,
	"DeCoy" <daleecoy@obfuscation.att.net> writes:
> To answer a number of replies:
> 
> [Footnote: somebody mentioned getting an account on Eisner - those are 
> "free", of course assuming that the system runs] 

That was me, and I thought I was being humorous as I am aware of the
nature of Eisner.  Just trying to stress that if I had something that
could speed the process of getting Eisner back up (even though I am
not and have never been a user) I would be glad to help.

bill

-- 
Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolves
billg999@cs.scranton.edu |  and a sheep voting on what's for dinner.
University of Scranton   |
Scranton, Pennsylvania   |         #include <std.disclaimer.h>   

I got curious about "Mylex" having never seen those contraptions. Google
lead me to a few goodies:

From ITRC:
> http://forums11.itrc.hp.com/service/forums/questionanswer.do?admit=109447626+1231109959807+28353475&threadId=1214006
##
 Andy Bustamante
Mar 24, 2008 16:46:15 GMT  9 pts 	
Odwillia,

The Mylex raid controllers were at one time an affordable entry level
raid solution. They had some quirks.

It's possible that your storage shelves lost power. This would cause the
controller to mark all disks as bad. One option, before resorting to
tape restore, is to power down the Alpha server, power down the disk,
restore power to the disk, wait for all disk to spin up, and restore
power to the Alpha.

You can also boot the Alphabios and run the Mylex configuration utility.
This will allow you to mark a device as good. This is a console
operation, preferably from a graphic console.

If these options don't work, your controller may have failed. If you've
saved the disk configuration to a floppy disk, a replacement controller
can be configured easily.
##

JF Mezei wrote:
> I got curious about "Mylex" having never seen those contraptions. Google
> lead me to a few goodies:
> 
> From ITRC:
>> http://forums11.itrc.hp.com/service/forums/questionanswer.do?admit=109447626+1231109959807+28353475&threadId=1214006
> ##
>  Andy Bustamante
> Mar 24, 2008 16:46:15 GMT  9 pts 	
> Odwillia,
> 
> The Mylex raid controllers were at one time an affordable entry level
> raid solution. They had some quirks.
> 

Had a Mylex in a AS-2100 once, a PITA.
After moving everything over to a HSZ40 shelf
everything was much easier to maintain. I do not
see why one would like to run a Mulex today...

JF Mezei wrote:
> I got curious about "Mylex" having never seen those contraptions. Google
> lead me to a few goodies:
> 
> From ITRC:
>> http://forums11.itrc.hp.com/service/forums/questionanswer.do?admit=109447626+1231109959807+28353475&threadId=1214006
> ##
>  Andy Bustamante
> Mar 24, 2008 16:46:15 GMT  9 pts 	
> Odwillia,
> 
> The Mylex raid controllers were at one time an affordable entry level
> raid solution. They had some quirks.
> 
> It's possible that your storage shelves lost power. This would cause the
> controller to mark all disks as bad. One option, before resorting to
> tape restore, is to power down the Alpha server, power down the disk,
> restore power to the disk, wait for all disk to spin up, and restore
> power to the Alpha.
> 
> You can also boot the Alphabios and run the Mylex configuration utility.
> This will allow you to mark a device as good. This is a console
> operation, preferably from a graphic console.
> 
> If these options don't work, your controller may have failed. If you've
> saved the disk configuration to a floppy disk, a replacement controller
> can be configured easily.
> ##
> 

This sounds vaguely familiar.  Were the Mylex controllers, by any 
chance, once known as "SWXCR"?

Richard B. Gilbert wrote:
> JF Mezei wrote:
>> I got curious about "Mylex" having never seen those contraptions. Google
>> lead me to a few goodies:
>>
>> From ITRC:
>>> http://forums11.itrc.hp.com/service/forums/questionanswer.do?admit=109447626+1231109959807+28353475&threadId=1214006 
>>>
>> ##
>>  Andy Bustamante
>> Mar 24, 2008 16:46:15 GMT  9 pts    
>> Odwillia,
>>
>> The Mylex raid controllers were at one time an affordable entry level
>> raid solution. They had some quirks.
>>
>> It's possible that your storage shelves lost power. This would cause the
>> controller to mark all disks as bad. One option, before resorting to
>> tape restore, is to power down the Alpha server, power down the disk,
>> restore power to the disk, wait for all disk to spin up, and restore
>> power to the Alpha.
>>
>> You can also boot the Alphabios and run the Mylex configuration utility.
>> This will allow you to mark a device as good. This is a console
>> operation, preferably from a graphic console.
>>
>> If these options don't work, your controller may have failed. If you've
>> saved the disk configuration to a floppy disk, a replacement controller
>> can be configured easily.
>> ##
>>
> 
> This sounds vaguely familiar.  Were the Mylex controllers, by any 
> chance, once known as "SWXCR"?

Sounds familiar...

To other replies:  Yes, I believe its a SWXCR
Thanks, JF - I've passed that info along.  It sounds applicable because 
there apparently WAS a power outage while Stephen was out of town.  Plus, 
the UPS battery needed to be replaced.

"Jan-Erik S�derholm" <jan-erik.soderholm@telia.com> wrote in message 
news:7tb8l.5387$U5.42850@newsb.telia.net...
> JF Mezei wrote:
>> I got curious about "Mylex" having never seen those contraptions. Google
>> lead me to a few goodies:
.........
> Had a Mylex in a AS-2100 once, a PITA.
> After moving everything over to a HSZ40 shelf
> everything was much easier to maintain. I do not
> see why one would like to run a Mylex today...

Volunteer-supported system, if it ain't broken don't fix it, and if it 
breaks, it MAY be easier to fix it, than to start over from backup tapes.

AND -- if we get the system working, then we can consider changing the disk 
configuration, and do so in an orderly fashion.

But, you understood that, I'm sure.

On Jan 4, 6:20=A0pm, "Richard B. Gilbert" <rgilber...@comcast.net>
wrote:

> This sounds vaguely familiar. =A0Were the Mylex controllers, by any
> chance, once known as "SWXCR"?

Yep. That's the one.  Finicky and a PITA as mentioned by others, but
about the only game in town for hardware RAID.  I've got a them
installed in AlphaSever 1200's and 800's but I haven't had them
running in about a year.

Jan-Erik S=F6derholm wrote:
> Richard B. Gilbert wrote:
>> JF Mezei wrote:
>>> I got curious about "Mylex" having never seen those contraptions. Goo=
gle
>>> lead me to a few goodies:
>>>
>>> From ITRC:
>>>> http://forums11.itrc.hp.com/service/forums/questionanswer.do?admit=3D=
109447626+1231109959807+28353475&threadId=3D1214006=20
>>>>
>>> ##
>>>  Andy Bustamante
>>> Mar 24, 2008 16:46:15 GMT  9 pts    Odwillia,
>>>
>>> The Mylex raid controllers were at one time an affordable entry level=

>>> raid solution. They had some quirks.
>>>
>>> It's possible that your storage shelves lost power. This would cause =
the
>>> controller to mark all disks as bad. One option, before resorting to
>>> tape restore, is to power down the Alpha server, power down the disk,=

>>> restore power to the disk, wait for all disk to spin up, and restore
>>> power to the Alpha.
>>>
>>> You can also boot the Alphabios and run the Mylex configuration utili=
ty.
>>> This will allow you to mark a device as good. This is a console
>>> operation, preferably from a graphic console.
>>>
>>> If these options don't work, your controller may have failed. If you'=
ve
>>> saved the disk configuration to a floppy disk, a replacement controll=
er
>>> can be configured easily.
>>> ##
>>>
>>
>> This sounds vaguely familiar.  Were the Mylex controllers, by any=20
>> chance, once known as "SWXCR"?

Yes


>=20
> Sounds familiar...



----== Posted via Pronews.Com - Unlimited-Unrestricted-Secure Usenet News==----
http://www.pronews.com The #1 Newsgroup Service in the World! >100,000 Newsgroups
---= - Total Privacy via Encryption =---

johnhreinhardt@yahoo.com wrote:
> On Jan 4, 6:20 pm, "Richard B. Gilbert" <rgilber...@comcast.net>
> wrote:
> 
>> This sounds vaguely familiar.  Were the Mylex controllers, by any
>> chance, once known as "SWXCR"?
> 
> Yep. That's the one.  Finicky and a PITA as mentioned by others, but
> about the only game in town for hardware RAID.  I've got a them
> installed in AlphaSever 1200's and 800's but I haven't had them
> running in about a year.

I didn't find the SWXCR to be all that difficult.  I picked up two or 
three of them on e-Bay.  They were dirt cheap.  I downloaded some 
utility software, installed the hardware and I was in business!

AIRC I installed one in an Alphaserver 2000 and one in an Alphaserver 
4100. I think I had one set up for RAID-5 and the other for RAID-1.

The directions for installation/setup were clear enough.  Once 
installed, they worked.  I don't think they are still running; the 
company was acquired, I was laid off, and I understand that the new 
owners used Novell for everything!  It was fun while it lasted!

On Jan 4, 5:13=A0pm, "DeCoy" <dalee...@obfuscation.att.net> wrote:

> I'll briefly quote Stephen here, in case somebody can spot something:
>
> <quote>
> In some of tries of the last experiment, the system console came alive
> and attempted to boot VMS.  You can see the last of the output in the
> screen shot athttp://www.Arnold.com/Photo_123108_002.jpg.  Note the
> message "SCSI drive at channel 2, target 1 dead".  The message about CPU
> 01 starting are not relevant.  Nothing else happens and there is no disk
> activity after that message.
>
> From this point I was able to ^P and give console commands. =A0Since
> storage is the problem, I entered "show disk". =A0The output is in the
> screen shot athttp://www.Arnold.com/Photo_123108_003.jpg. =A0DRA0-5 are
> drives defined on the seven-member RAID set in the built-in shelf, on
> channels 1-2 of the Mylex controller. =A0DRA6 is on the six-member RAID
> set in the external gray storage shelf. =A0(We also have the FTP
> repository on large ATA drive connected via an ATA-SCSI adpater: DKA0.)
> <end of quote>
>

Just for clarification of the configuration - you've got a DS20 with a
Mylex (swxcr) scsi RAID controller using 2 channels on a split SCSI
bus on the stack of 7 drive slots going up the right side.  These 7
drives are 9GB bricks in a RAID-5 and from that you've sliced it into
6 logical drives DRA0-5.  The drive in slot #1 on the second bus has
failed so the swxcr has marked DRA0-5 as "DEGRADED".  Then the 3rd
channel of the controller is used for DRA6 on an external shelf.

I've seen the "%DRA, drives=3D0, optimal =3D 4294967290, degraded =3D 6,
failed =3D 0 " before.  I think it's a combination of a firmware bug and
a slightly confused controller.   I wish I could remember exactly what
I did to fix it but it was a few years ago when I had the problem.

I would assume that you've tried replacing the bad disk mentioned with
another?  The swxcr may not automatically do a rebuild.  You may have
to go into the swxcrmgr utility and tell the controller that the drive
has been replaced, mark it as good and then it to rebuild the array.

Did the external shelf become disconnected or lose power?  It doesn't
look like the swxcr sees it at the moment. That might be enough to
hang it, though if that's the case is should show as failed.  I'm not
sure why there is no mention of it in the startup messages.  It may be
irrelevant if the DRA0-5 array is rebuilt.

I know you're just passing on info second hand, but a little more info
on what's been tried and failed and what the failure messages/results
were would help.

DeCoy wrote:
> Thanks, George.  The problem appears to be storage-related, perhaps with the 
> RAID array on the Mylex controller, and perhaps with either controller 
> hardware or controller configuration.
> 
> Expertise in diagnosing (and perhaps fixing) Mylex controller symptoms would 
> be initially useful.

I won't be of much help - my experience is with the HSJ/HSZ/HSG controller
series.

I did leave voice mail for Steve offering my services, but you folks
will probably do better diagnosing it remotely than me trying to get
involved.  Let me know, though, if I can do something, even if it's
just getting him some spare parts.

Coincidentally, the reason I am not using the DS20 that's in my garage
is that the Mylex (KZPBC?) controller failed when I was trying to configure
it and I have not yet sprung for a replacement or stuffed in a non-raid
SCSI card.

I know of others around here who have used the Mylex controller and
have encountered some of its quirks.  I seem to remember helping someone
on the research side of things restore a backup of what was at the time
a large (30GB) raid volume that was lost due to Mylex controller issues,
or perhaps due to not noticing that a raid disk had failed until a
second failure made recovery impossible.

-- 
George Cornelius                 cornelius at mayo dot edu

DeCoy wrote:
> To answer a number of replies:
> 
> I have suggested that the easiest and most straightforward solution is to 
> get the Mylex-driven RAID 5 array to work.
> 
> We have spare 9GB drives for that.  My suggestion was that Stephen, with 
> whatever support and advice are available, work on that FIRST, to see if 
> it's viable (given hardware, software, and peopleware).
> 
> If we can't do that, then we'll explore alternatives for replacing the whole 
> disk setup.  We may have some in hand, but offers are appreciated.  We would 
> certainly go to bigger drives, at that point.
> 
> I'll briefly quote Stephen here, in case somebody can spot something:
> 
> <quote>
> In some of tries of the last experiment, the system console came alive
> and attempted to boot VMS.  You can see the last of the output in the
> screen shot at http://www.Arnold.com/Photo_123108_002.jpg.  Note the
> message "SCSI drive at channel 2, target 1 dead".  The message about CPU
> 01 starting are not relevant.  Nothing else happens and there is no disk
> activity after that message.

So the raid set is working long enough to bring in the kernel but
VMS cannot continue from there?  I assume that is true even with a
conversational boot?

Is he able to run the SWXCR utility from floppy?  It requires
a video monitor, I believe, and may require switching to the
NT-compatible BIOS (AlphaBIOS?).

What does booting from CD give him?

Are dra0-6 the virtual units, or the physical raid set members?

[Just some things that come to mind looking at the screen captures].

-- 
George Cornelius                  cornelius at mayo dot edu

> From this point I was able to ^P and give console commands.  Since
> storage is the problem, I entered "show disk".  The output is in the
> screen shot at http://www.Arnold.com/Photo_123108_003.jpg.  DRA0-5 are
> drives defined on the seven-member RAID set in the built-in shelf, on
> channels 1-2 of the Mylex controller.  DRA6 is on the six-member RAID
> set in the external gray storage shelf.  (We also have the FTP
> repository on large ATA drive connected via an ATA-SCSI adpater: DKA0.)
> <end of quote>
> 
> 
> 
> [Footnote: somebody mentioned getting an account on Eisner - those are 
> "free", of course assuming that the system runs] 

DeCoy wrote:
> In some of tries of the last experiment, the system console came alive
> and attempted to boot VMS.  You can see the last of the output in the
> screen shot at http://www.Arnold.com/Photo_123108_002.jpg.  Note the
> message "SCSI drive at channel 2, target 1 dead".  The message about CPU
> 01 starting are not relevant.  Nothing else happens and there is no disk
> activity after that message.

I'm guessing the fact that the controller thinks it has no drives, but 4 
billion optimal drives (in addition to 6 degraded ones) is not a good 
sign...

I hated hated hated the SWXCR.  When it worked it did a fine job, but it 
was the fear of this kind of spurious problem, just about impossible to 
recover from, that made me happy to stop using the thing.  Not to 
mention the annoyance of having to take down VMS and boot SWXCRMGR to do 
any work on it.

Any chance he's got a good backup of the config on floppy and can reload 
it into a replacement controller?

Duncan

G Cornelius wrote:
> DeCoy wrote:
>> Thanks, George.  The problem appears to be storage-related, perhaps with the 
>> RAID array on the Mylex controller, and perhaps with either controller 
>> hardware or controller configuration.
>>
>> Expertise in diagnosing (and perhaps fixing) Mylex controller symptoms would 
>> be initially useful.
> 
> I won't be of much help - my experience is with the HSJ/HSZ/HSG controller
> series.
> 
> I did leave voice mail for Steve offering my services, but you folks
> will probably do better diagnosing it remotely than me trying to get
> involved.  Let me know, though, if I can do something, even if it's
> just getting him some spare parts.
> 
> Coincidentally, the reason I am not using the DS20 that's in my garage
> is that the Mylex (KZPBC?) controller failed when I was trying to configure
> it and I have not yet sprung for a replacement or stuffed in a non-raid
> SCSI card.
> 
> I know of others around here who have used the Mylex controller and
> have encountered some of its quirks.  I seem to remember helping someone
> on the research side of things restore a backup of what was at the time
> a large (30GB) raid volume that was lost due to Mylex controller issues,
> or perhaps due to not noticing that a raid disk had failed until a
> second failure made recovery impossible.
> 

It seems to me that it's a SYS$MANGLER's JOB to notice things like 
failing disks.  I had a batch job called "MORNING_CHECK" that ran every 
day at  07:30.  It compared the output of "SHOW ERROR" with the output 
from yesterday.  It checked log files for errors ("-E-" and -F-"), etc, 
etc.  If it found something that looked like a problem I was notified by 
a text message to my pager.  This gave me time to work on the problem 
before it turned into a crisis!

A failed disk was not allowed to become a problem!  I would swap it out 
with a spare and call DEC/Compaq/HP to pick up the dear departed and 
bring me a replacement drive.

In fact, thanks to MORNING_CHECK, I usually found disks that were 
developing problems before the problems developed fully.  One error was 
allowed but when a disk started logging multiple errors, I swapped it 
out with a spare and called for a replacement.  The same guy who fetched 
replacements for field service would fetch me a new one and I gave him 
the dear departed!

I'll pass along the questions below, but answers may not be returned.

Please note that I am 400 miles from the system, and have never physically 
touched it, and that various aspects are handled by various volunteers. 
That's one of the "interesting" aspects of Eisner.  Another is that people 
do have "day jobs".

To (briefly) answer another posting - yes, there are various automatic 
system utilities that try to warn us of impending problems.  And there's a 
UPS (unfortunately, it turns out, with a bad battery).  We don't have a 
utility that warns of an upcoming power failure in the neighborhood.


<johnhreinhardt@yahoo.com> wrote in message 
news:7172eb31-0912-4bd6-b440-2955797ba148@u18g2000pro.googlegroups.com...


Just for clarification of the configuration - you've got a DS20 with a
Mylex (swxcr) scsi RAID controller using 2 channels on a split SCSI
bus on the stack of 7 drive slots going up the right side.  These 7
drives are 9GB bricks in a RAID-5 and from that you've sliced it into
6 logical drives DRA0-5.  The drive in slot #1 on the second bus has
failed so the swxcr has marked DRA0-5 as "DEGRADED".  Then the 3rd
channel of the controller is used for DRA6 on an external shelf.

I've seen the "%DRA, drives=0, optimal = 4294967290, degraded = 6,
failed = 0 " before.  I think it's a combination of a firmware bug and
a slightly confused controller.   I wish I could remember exactly what
I did to fix it but it was a few years ago when I had the problem.

I would assume that you've tried replacing the bad disk mentioned with
another?  The swxcr may not automatically do a rebuild.  You may have
to go into the swxcrmgr utility and tell the controller that the drive
has been replaced, mark it as good and then it to rebuild the array.

Did the external shelf become disconnected or lose power?  It doesn't
look like the swxcr sees it at the moment. That might be enough to
hang it, though if that's the case is should show as failed.  I'm not
sure why there is no mention of it in the startup messages.  It may be
irrelevant if the DRA0-5 array is rebuilt.

I know you're just passing on info second hand, but a little more info
on what's been tried and failed and what the failure messages/results
were would help.

Any further news on poor little old Eisner?  Anything I can help with? 
My memories of how to beat on a SWXCR are fading, but I might be able to 
be of some use.  Or should we be looking for spare parts on ebay, or...?

Duncan

"Duncan Brown" <brown_du@eisner.decus.org> wrote in message 
news:RdydnVL-PpHNmfnUnZ2dnUVZ_oHinZ2d@speakeasy.net...
> Any further news on poor little old Eisner?  Anything I can help with? My 
> memories of how to beat on a SWXCR are fading, but I might be able to be 
> of some use.  Or should we be looking for spare parts on ebay, or...?
>
> Duncan

No additional info from Stephen.  I've sent him everything "significant" 
that has appeared in comp.os.vms, including offers of personal, electronic, 
and monetary support. 

I really miss Eisner...

I wonder if, compared to a SWXCR, just letting drives die every once in 
a blue moon and replacing them and restoring them from backups isn't a 
better uptime strategy... that's the decision I came to on my system!

Duncan

In article <JNqdnW9V0b1J3_rUnZ2dnUVZ_uGdnZ2d@speakeasy.net>, Duncan Brown <brown_du@eisner.decus.org> writes:
>I really miss Eisner...
>
>I wonder if, compared to a SWXCR, just letting drives die every once in 
>a blue moon and replacing them and restoring them from backups isn't a 
>better uptime strategy... that's the decision I came to on my system!

For Eisner, I'd toss a vote in the ballot box for HBVS.  Most of the 
hardware is already there.  I've experienced several sites using the
SWXCR and, to me, the down-time to boot into ARC, run the rcu to fix
the problem, and then reboot SRM and VMS is not worth it in most, if
not all, of the cases I've been privy.

-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

  "Well my son, life is like a beanstalk, isn't it?"

VAXman- @SendSpamHere.ORG wrote:

> For Eisner, I'd toss a vote in the ballot box for HBVS.  

If they used raid-5 to bind multiple small disks into a big one, why not
just buy 1 modern disk and put it all on one disk and just back it up ?
It would be a lot simpler and less chance of failure.

Duncan Brown wrote:
> I really miss Eisner...
> 
> I wonder if, compared to a SWXCR, just letting drives die every once in 
> a blue moon and replacing them and restoring them from backups isn't a 
> better uptime strategy... that's the decision I came to on my system!
> 

It works.  IFF your backups are made at sufficiently frequent intervals 
that you don't lose anything important AND failures are not frequent!

If I needed 99.99% uptime, such a strategy would be risking my job!  And 
I don't think I would be using technology that's now at least ten years 
old!  The SWXCR is AT LEAST that old.

EISNER being a low/no budget volunteer based service, you have to expect 
some downtime.

Richard B. Gilbert wrote:
> If I needed 99.99% uptime, such a strategy would be risking my job!  And 
> I don't think I would be using technology that's now at least ten years 
> old!  The SWXCR is AT LEAST that old.
> 
> EISNER being a low/no budget volunteer based service, you have to expect 
> some downtime.
> 

I was using (and learning to hate) SWXCRs in AlphaServer 2100 systems 
back in 1996, so that's 12-13 years at least.

I expect some downtime with Eisner, which is why the lack of RAID would 
seem to be perfectly acceptable.  It certainly takes a lot less than 20 
days to toss another cheap big disk in and restore from backups.  I 
wasn't recommending that strategy for a business system, no!

Duncan

JF Mezei wrote:
> VAXman- @SendSpamHere.ORG wrote:
> 
>> For Eisner, I'd toss a vote in the ballot box for HBVS.  
> 
> If they used raid-5 to bind multiple small disks into a big one, why not
> just buy 1 modern disk and put it all on one disk and just back it up ?
> It would be a lot simpler and less chance of failure.
> 

The question is never WHETHER a disk drive will fail.  The drive WILL 
fail; the only question is when!

When that one big disk fails, the system is down.  Again!

ISTR that you can have a "hot spare" in a RAID-5 set with a hardware 
RAID controller and that drive will replace the first member of the set 
to fail.  If the SYS$MANGLER is paying attention, he should notice the 
failure no later than the first work day following the failure.

I made it a point both to check my error counters and log files and ALSO 
to visit the computer room at least once a day to look for yellow lights 
on the drives or failed power supplies, listen for "that loud scraping 
sound", look for things turned off that should be on, etc.

I found a few problems that way and was generally able to prevent 
problems from turning into crises!

Duncan Brown wrote:
> Richard B. Gilbert wrote:
>> If I needed 99.99% uptime, such a strategy would be risking my job!  
>> And I don't think I would be using technology that's now at least ten 
>> years old!  The SWXCR is AT LEAST that old.
>>
>> EISNER being a low/no budget volunteer based service, you have to 
>> expect some downtime.
>>
> 
> I was using (and learning to hate) SWXCRs in AlphaServer 2100 systems 
> back in 1996, so that's 12-13 years at least.
> 
> I expect some downtime with Eisner, which is why the lack of RAID would 
> seem to be perfectly acceptable.  It certainly takes a lot less than 20 
> days to toss another cheap big disk in and restore from backups.  I 
> wasn't recommending that strategy for a business system, no!
> 

I encountered SWXCRs ca. 2002 although I had heard of them years before 
that.  I bought two of them, used, for use with an Alphaserver 2000 and 
and an Alphaserver 4100.  It was "cheap RAID" and I needed that.  I 
think I paid about $35/each.  One was configured to do RAID-5 and the 
other did a couple of RAID-1 sets.  I used them for about two years and 
don't recall having any problems with them.

As for twenty days downtime, it's not good but it may be the best 
service available.  When a service is operated out of someone's home you 
learn to expect power failures, downtime when something fails and the 
owner is 3,000 miles away on a business trip or vacation, etc.  He may 
have died and his family hasn't a clue. . . .

If you want 24x365, you pay for the service.  There are companies that 
run data centers for profit and will host your machines/services for a 
moderately outrageous monthly fee.  They tend to have things like power 
feeds from two different substations, emergency generators, redundant 
air conditioners, T1/T3 service from two or more "points of presence", etc.

--
draco vulgaris

I do not want to sound picky, but it's now been over three weeks. This
is the point where a little information as a common courtesy would be
appropriate. I fully understand that the person running this is doing
so on a voluntary basis, but this begs the point, that perhaps it is
time to ask others to help, either in terms of supply parts, or
standing in with the necessary expertise. If a volunteer, needs help,
then they should say so.

Equally, some accurate information, on what is needed to be done, when
it is planned for, what problems have been encountered, would be
beneficial, not only to those that are patiently waiting, but also in
digging out those, that may be able to provide some assistance. ( I
have several spare drives, and a number of disk controllers, which
might include one of the model in question. - although I am half way
around the world....)

Peter

On 2009-01-12, Mister Q <pquodling@gmail.com> wrote:
> I do not want to sound picky, but it's now been over three weeks. This
> is the point where a little information as a common courtesy would be
> appropriate. I fully understand that the person running this is doing
> so on a voluntary basis, but this begs the point, that perhaps it is
> time to ask others to help, either in terms of supply parts, or
> standing in with the necessary expertise. If a volunteer, needs help,
> then they should say so.
>

A couple of days ago, Dale asked for an update on a Eisner mailing list
which has been setup. (No, I don't know how to join the mailing list;
I asked Steve for an update and was added to the list instead. :-))

I haven't seen any response as yet, but hopefully one will be made.

Simon.

-- 
Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP
Microsoft: Bringing you 1980's technology to a 21st century world
(Note: Email address not currently working)

One could simply say: Just tell people they are welcome on comp.os.vms .

However, due to upcoming events, I would strongly suggest that some
urgency be put on restoring eisner's service, even if it means dropping
that unreliable disk array and just putting everything on a single 160
gig drive or whatever they can get.

Eisner may find itself with a whole bunch of valuable users that will
make ITRC look like a kids playpen.

"Mister Q" <pquodling@gmail.com> wrote in message 
news:22f49a72-093f-4a96-b48a-8c129d307e0d@t39g2000prh.googlegroups.com...
>I do not want to sound picky, but it's now been over three weeks. This
> is the point where a little information as a common courtesy would be
> appropriate. I fully understand that the person running this is doing
> so on a voluntary basis, but this begs the point, that perhaps it is
> time to ask others to help, either in terms of supply parts, or
> standing in with the necessary expertise. If a volunteer, needs help,
> then they should say so.
>
> Equally, some accurate information, on what is needed to be done, when
> it is planned for, what problems have been encountered, would be
> beneficial, not only to those that are patiently waiting, but also in
> digging out those, that may be able to provide some assistance. ( I
> have several spare drives, and a number of disk controllers, which
> might include one of the model in question. - although I am half way
> around the world....)
>
> Peter

Your concern, and your offers of help, are appreciated.  I have posted here 
***ALL*** of the information that I have.

....which is, as you point out, damned little.

Unless/until Stephen asks for help, it's not practical to provide help.

I have attempted to encourage Stephen to provide an update on 
status/prognosis.  If I had received any answer, I would have posted it 
here.

Here is Stephen's contact information.  Perhaps someone else can find out 
more than I can.

"Steve"   Stephen L. Arnold, Ph.D., President, Arnold Consulting, Inc.
Address   2530 Targhee Street, Fitchburg, Wisconsin  53711-5491  U.S.A.
Telephone +1 608 278 7700               Facsimile +1 608 278 7701
Internet  Stephen.L.Arnold@Arnold.com   http://WWW.Arnold.com
Arnold� is a registered trademark and service mark of Arnold Consulting, 
Inc. 

Zzzzzzzzzzzzzzzzzzz

Wake me when something happens. I am sure that I have a SWXCR in a box
in the Garage if really needed - I can possibly spring out a spare
disk drive, as well...

Peter

On Jan 13, 6:31=A0am, Mister Q <pquodl...@gmail.com> wrote:
> I do not want to sound picky, but it's now been over three weeks. This
> is the point where a little information as a common courtesy would be
> appropriate. I fully understand that the person running this is doing
> so on a voluntary basis, but this begs the point, that perhaps it is
> time to ask others to help, either in terms of supply parts, or
> standing in with the necessary expertise. If a volunteer, needs help,
> then they should say so.
>
> Equally, some accurate information, on what is needed to be done, when
> it is planned for, what problems have been encountered, would be
> beneficial, not only to those that are patiently waiting, but also in
> digging out those, that may be able to provide some assistance. ( I
> have several spare drives, and a number of disk controllers, which
> might include one of the model in question. - although I am half way
> around the world....)
>
> Peter

No news, just updating the subject line :-(

Duncan

I share your concern, and if I had any information at all, I would have 
posted it.


"Mister Q" <pquodling@gmail.com> wrote in message 
news:4a2ea9b7-6023-41a4-be2c-a44278721d8c@r36g2000prf.googlegroups.com...
Zzzzzzzzzzzzzzzzzzz

Wake me when something happens. I am sure that I have a SWXCR in a box
in the Garage if really needed - I can possibly spring out a spare
disk drive, as well...

Peter

Just another long time user missing EISNER... hoping for an update.

<htmlworx@gmail.com> wrote in message 
news:18633a73-5eb5-4cc1-9526-e017bde4c45e@f11g2000vbf.googlegroups.com...
> Just another long time user missing EISNER... hoping for an update.
>

Wish I could provide an update.

Anybody here who is near Fitchburg (Madison) Wisconsin, and would consider 
going to ask Stephen what he needs? 

On Jan 22, 10:14=A0pm, "DeCoy" <dalee...@obfuscation.att.net> wrote:
> <htmlw...@gmail.com> wrote in message
>
> news:18633a73-5eb5-4cc1-9526-e017bde4c45e@f11g2000vbf.googlegroups.com...
>
> > Just another long time user missing EISNER... hoping for an update.
>
> Wish I could provide an update.
>
> Anybody here who is near Fitchburg (Madison) Wisconsin, and would conside=
r
> going to ask Stephen what he needs?

I suspect he'll say 'time'.
Looks like he's busy in the local government.

http://www.arnold.us/
http://www.arnold.us/arnold/alder/history.htmlx

That first web page indicates 'last updated 20 January 2009'
A sign of life ?!

Hein.
(Who has been booting his own system(s) much more often these weeks!)

Just an idea...

Send the backups to someone else. Someone else mounts them on a disk
that is NFS mounted by the original site. Might be slow but it least it
would be available.

(or is NFS incompatible with RMS indexed files which Notes uses ?)

The concept of having a backup site runned by another volunteer might
work well  here. (Especially when 8.4 with IP-Clustering becomes available)

Is this based on VAX or Alpha ?

How much disk space is needed for this service ?

In article <0007343a$0$2081$c3e8da3@news.astraweb.com>, JF Mezei <jfmezei.spamnot@vaxination.ca> writes:
>Just an idea...
>
>Send the backups to someone else. Someone else mounts them on a disk
>that is NFS mounted by the original site. Might be slow but it least it
>would be available.
>
>(or is NFS incompatible with RMS indexed files which Notes uses ?)
>
>The concept of having a backup site runned by another volunteer might
>work well  here. (Especially when 8.4 with IP-Clustering becomes available)
>
>Is this based on VAX or Alpha ?

Eisner is/was an Alpha DS20, IIRC.  I have a full-blown DS10 currently
not in use for anything important.  I could, if I can read the backups
to restore, configure a substitute for those needing an Eisner fix.


>How much disk space is needed for this service ?

Hmm...  I don't recall the actual storage capabilities of Eisner.
-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

  "Well my son, life is like a beanstalk, isn't it?"

VAXman- @SendSpamHere.ORG wrote:
>  I have a full-blown DS10 currently
> not in use for anything important. 


What is the model number ? I had never heard about inflatable Alphas.

I take it that it can be stored in a small drawer when deflated ?

On Jan 23, 10:12=A0am, JF Mezei <jfmezei.spam...@vaxination.ca> wrote:
> VAXman- @SendSpamHere.ORG wrote:
> > =A0I have a full-blown DS10 currently
> > not in use for anything important.
>
> What is the model number ? I had never heard about inflatable Alphas.

Cute. :-).

It was a DS20. 2-cpu.
The main storage was on an old  KZPCA (SWXCR), with 4GB spindles I
believe.
Those things could only present a logical drives up to 36 GB each.
Should not be a big deal.

As I seem to recall from the Notes article suggesting funding Steve
Arnolds main cost concern was the network connection he used, and a
little about the power for an always-on system.

Hein.

In article <81c6f9bb-bf7b-418a-a4c7-a4ec9f861572@z27g2000prd.googlegroups.com>, Hein RMS van den Heuvel <heinvandenheuvel@gmail.com> writes:
>On Jan 23, 10:12=A0am, JF Mezei <jfmezei.spam...@vaxination.ca> wrote:
>> VAXman- @SendSpamHere.ORG wrote:
>> > =A0I have a full-blown DS10 currently
>> > not in use for anything important.
>>
>> What is the model number ? I had never heard about inflatable Alphas.
>
>Cute. :-).
>
>It was a DS20. 2-cpu.
>The main storage was on an old  KZPCA (SWXCR), with 4GB spindles I
>believe.
>Those things could only present a logical drives up to 36 GB each.
>Should not be a big deal.
>
>As I seem to recall from the Notes article suggesting funding Steve
>Arnolds main cost concern was the network connection he used, and a
>little about the power for an always-on system.

....and that was a T1 service.  I believe it supported more than just
DECUServe.

-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

  "Well my son, life is like a beanstalk, isn't it?"

"Hein RMS van den Heuvel" <heinvandenheuvel@gmail.com> wrote in message 
news:f6087bc7-4e54-44ea-a0d0-6a53b51c0566@e18g2000vbe.googlegroups.com...
On Jan 22, 10:14 pm, "DeCoy" <dalee...@obfuscation.att.net> wrote:
> <htmlw...@gmail.com> wrote in message
>
> news:18633a73-5eb5-4cc1-9526-e017bde4c45e@f11g2000vbf.googlegroups.com...
>
> > Just another long time user missing EISNER... hoping for an update.
>
> Wish I could provide an update.
>
> Anybody here who is near Fitchburg (Madison) Wisconsin, and would consider
> going to ask Stephen what he needs?

I suspect he'll say 'time'.
Looks like he's busy in the local government.

http://www.arnold.us/
http://www.arnold.us/arnold/alder/history.htmlx

That first web page indicates 'last updated 20 January 2009'
A sign of life ?!

Hein.
(Who has been booting his own system(s) much more often these weeks!)

Yes, if Stephen is still involved with the Executive Office of the 
President, he's undoubtedly very busy there.

I've offered Stephen my help when I talked to him on the phone, 
including being willing to drive up there (about 2 hours away from me) 
to work on it.  Apparently others in Wisconsin have made similar offers. 
  I've followed up with some SWXCR tips in email, based on my memories 
and notes from when I used to use them.  It does sound like his lack of 
free time is one of the biggest problems, but we're now at 30 days and 
still counting...

Duncan

Duncan Brown wrote:

> and notes from when I used to use them.  It does sound like his lack of 
> free time is one of the biggest problems, but we're now at 30 days and 
> still counting...


I think all of you should demand your money back because of the downtime
 :-) :-) ;-) :-) :-) ;-) :-)

JF Mezei wrote:
> Duncan Brown wrote:
> 
>> and notes from when I used to use them.  It does sound like his lack of 
>> free time is one of the biggest problems, but we're now at 30 days and 
>> still counting...
> 
> 
> I think all of you should demand your money back because of the downtime
>  :-) :-) ;-) :-) :-) ;-) :-)

I just wish somebody would think of a solution that does not involve 
flooding c.o.v. with complaints!  The current "solution" is obviously 
not effective.  How about trying something else?

I don't see anyone else volunteering to install T1 service and host Eisner!

I have not been a DECUS member since the "blowup" back in the late 
nineties!  It looks as if I haven't been missing much!

Richard B. Gilbert wrote:

> I don't see anyone else volunteering to install T1 service and host Eisner!

OK, I guess it is time to step up to the plate.   If it becomes
necessary for DECUServe to move, I will give it a new home.

I have a T1 at my home/office.   I know it is not very fast by today's
standards, but the time-to-repair is great.   Both Covad (who I write
the checks to) and AT&T (who provisioned the local loop) are really
good about fixing the occasional outage.   The sole exception was on
Christmas day, when it took AT&T eleven hours to dispatch someone to
fix the line.

I am willing to make space, power, a fixed IP address, and DNS service
available for Eisner.   I will perform regular backups and whatever
maintenance is necessary to keep it running.   All I ask is that
whoever currently performs system/DECUServe management (presumably
remotely) will continue to do so.

Why am I willing to do this?   Well...

Both my wife (Sharon Johnson) and I were very early moderators
on the system, long before it was named Eisner.   Dan Eisner recruited
us at the 1987 Spring DECUS symposium in Nashville.   It was because
of this involvement that we grew to appreciate each other, and
eventually get married.   Dan was the first person we told about our
engagement, just before he died.   Dan was a good friend, and I miss
him very much.

This is just my way of trying to repay the huge debt that I owe to
Dan Eisner and DECUServe.

OK, I've made the offer.   Does DECUServe management wish to take
me up on it?

Alan Frisbie   (Remove _REMOVE to reply)
Flying Disk Systems, Inc.
Los Angeles, CA

On Jan 29, 7:21=A0pm, Alan Frisbie <Usenet02_Rem...@Flying-Disk.com>
wrote:
> Richard B. Gilbert wrote:
> > I don't see anyone else volunteering to install T1 service and host Eis=
ner!
>
> OK, I guess it is time to step up to the plate. =A0 If it becomes
> necessary for DECUServe to move, I will give it a new home.

Nice. Very nice.

Hein.

In article <zZKdnbKkAvOX1R_UnZ2dnUVZ_jWWnZ2d@supernews.com>, Alan Frisbie <Usenet02_Remove@Flying-Disk.com> writes:
>Richard B. Gilbert wrote:
>
>> I don't see anyone else volunteering to install T1 service and host Eisner!
>
>OK, I guess it is time to step up to the plate.   If it becomes
>necessary for DECUServe to move, I will give it a new home.
>
>I have a T1 at my home/office.   I know it is not very fast by today's
>standards, but the time-to-repair is great.   Both Covad (who I write
>the checks to) and AT&T (who provisioned the local loop) are really
>good about fixing the occasional outage.   The sole exception was on
>Christmas day, when it took AT&T eleven hours to dispatch someone to
>fix the line.
>
>I am willing to make space, power, a fixed IP address, and DNS service
>available for Eisner.   I will perform regular backups and whatever
>maintenance is necessary to keep it running.   All I ask is that
>whoever currently performs system/DECUServe management (presumably
>remotely) will continue to do so.
>
>Why am I willing to do this?   Well...
>
>Both my wife (Sharon Johnson) and I were very early moderators
>on the system, long before it was named Eisner.   Dan Eisner recruited
>us at the 1987 Spring DECUS symposium in Nashville.   It was because
>of this involvement that we grew to appreciate each other, and
>eventually get married.   Dan was the first person we told about our
>engagement, just before he died.   Dan was a good friend, and I miss
>him very much.
>
>This is just my way of trying to repay the huge debt that I owe to
>Dan Eisner and DECUServe.
>
>OK, I've made the offer.   Does DECUServe management wish to take
>me up on it?
>
>Alan Frisbie   (Remove _REMOVE to reply)
>Flying Disk Systems, Inc.
>Los Angeles, CA


I thought about that too Alan.  I have about 20 times your bandwidth but
if you want the duties, you're welcome to it and I know you'd do a good
job too.


-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

  "Well my son, life is like a beanstalk, isn't it?"

"Alan Frisbie" <Usenet02_Remove@Flying-Disk.com> wrote in message 
news:zZKdnbKkAvOX1R_UnZ2dnUVZ_jWWnZ2d@supernews.com...
>
> OK, I guess it is time to step up to the plate.   If it becomes
> necessary for DECUServe to move, I will give it a new home.
>
.....
>
> I am willing to make space, power, a fixed IP address, and DNS service
> available for Eisner.   I will perform regular backups and whatever
> maintenance is necessary to keep it running.   All I ask is that
> whoever currently performs system/DECUServe management (presumably
> remotely) will continue to do so.
............
>
> OK, I've made the offer.   Does DECUServe management wish to take
> me up on it?
>
> Alan Frisbie   (Remove _REMOVE to reply)
> Flying Disk Systems, Inc.
> Los Angeles, CA

Thanks, Alan.  I guess I'm "management" at the moment.  I will pass along 
the offer to Steve, and if he agrees, we'll make it happen some way. 
Although I'm "management", I believe the decision is really up to Steve.

In that event, I can assure you that "remote system management" will 
continue as in the past - or can be modified if you wish.

As I said, I'll pass along the (full) offer by email.  However, if you're 
willing, perhaps you would consider calling Steve personally.  I'm sure you 
know each other.
["Steve"   Stephen L. Arnold,  Telephone +1 608 278 7700]

And thanks again for the exceedingly generous offer.

Dale E. Coy 

VAXman- @SendSpamHere.ORG wrote:
> In article <zZKdnbKkAvOX1R_UnZ2dnUVZ_jWWnZ2d@supernews.com>, Alan Frisbie <Usenet02_Remove@Flying-Disk.com> writes:
>> Richard B. Gilbert wrote:
>>
>>> I don't see anyone else volunteering to install T1 service and host Eisner!
>> OK, I guess it is time to step up to the plate.   If it becomes
>> necessary for DECUServe to move, I will give it a new home.

> 
> I thought about that too Alan.  I have about 20 times your bandwidth but
> if you want the duties, you're welcome to it and I know you'd do a good
> job too.
> 
> 

Yes, thanks Alan. I was also wrestling with whether or not to volunteer 
to host it (easy to say after someone else has already volunteered I guess).

I have lots of spare VMS capable hardware here and a 100Mbit pipe, but 
I'm not sure what sort of bureaucracy I'd have to go through to get my 
employer (a university) to approve it. I have to admit I've kind of 
drifted away from Emcompassserve lately anyway(not from VMS though), so 
I'm not personally suffering from Eisner being down.

Also, I'm in Canada and hosting it here would go against the current 
protectionist spirit of the US Congress :-)

In article <49833245$1@flight>, Malcolm Dunnett <nothome@spammers.are.scum> writes:
>VAXman- @SendSpamHere.ORG wrote:
>> In article <zZKdnbKkAvOX1R_UnZ2dnUVZ_jWWnZ2d@supernews.com>, Alan Frisbie <Usenet02_Remove@Flying-Disk.com> writes:
>>> Richard B. Gilbert wrote:
>>>
>>>> I don't see anyone else volunteering to install T1 service and host Eisner!
>>> OK, I guess it is time to step up to the plate.   If it becomes
>>> necessary for DECUServe to move, I will give it a new home.
>
>> 
>> I thought about that too Alan.  I have about 20 times your bandwidth but
>> if you want the duties, you're welcome to it and I know you'd do a good
>> job too.
>> 
>> 
>
>Yes, thanks Alan. I was also wrestling with whether or not to volunteer 
>to host it (easy to say after someone else has already volunteered I guess).
>
>I have lots of spare VMS capable hardware here and a 100Mbit pipe, but 
>I'm not sure what sort of bureaucracy I'd have to go through to get my 
>employer (a university) to approve it. I have to admit I've kind of 
>drifted away from Emcompassserve lately anyway(not from VMS though), so 
>I'm not personally suffering from Eisner being down.
>
>Also, I'm in Canada and hosting it here would go against the current 
>protectionist spirit of the US Congress :-)

I believe the h/w is DECUServer/EncompasServe/name-of-the-weekServe.  I'd
assume the h/w and any relevant backups, etc. would be shipped out to any
party that would take over its hosting.


-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

  "Well my son, life is like a beanstalk, isn't it?"

VAXman- @SendSpamHere.ORG wrote:

> I believe the h/w is DECUServer/EncompasServe/name-of-the-weekServe.  I'd
> assume the h/w and any relevant backups, etc. would be shipped out to any
> party that would take over its hosting.

The costs of shipping can be horrendous for computers.

However, shipping of some peripherals (do do backups etc) might have to
be involved.

However, bear in mind that if the current system hoster is extremely
busy, he may not have the time to "pack" decuserve and ship it.

VAXman wrote:

>> I am willing to make space, power, a fixed IP address, and DNS service
>> available for Eisner.

> I thought about that too Alan.  I have about 20 times your bandwidth but
> if you want the duties, you're welcome to it and I know you'd do a good
> job too.

I didn't say I *wanted* the job.   :-)   :-)   :-)
However, there is this old-fashioned thing called "duty", which
I still believe in.   Fortunately, Sharon was very understanding
when I told her that DECUServe might be coming to live with us.  :-)

I think the users would be better served by a bigger pipe than
mine, but I can't get it here -- at least for a price mere mortals
can afford.   :-)   It hurts to pay $389/month for the T1, but the
next lower price is $156/month for IDSL (144kbps).   It sucks
being 22,000 feet from the Telco CO.   I looked into wireless, but
there is a large hill in the way.   :-(   Business-class cable is
not available.

If my offer spurs someone with better facilities to make a similar
offer, that would be wonderful and I will bow to them.   If not,
or if a backup site is desired, my offer stands.

Alan Frisbie

How difficult would it be to distribute eisner amongst multiple volunteers ?

Once IP-cluster is available, wouldn't it be possible to volume-shadow
the disk containing the notes files ? And then make the host name
resolve to multiple IPs for load balancing, hence requiring less
bandwidth from each volunteer ?

Apart from volume shadowing, are there any other means to have notes
files distributed amongst multiple sites ?

JF Mezei wrote:
> VAXman- @SendSpamHere.ORG wrote:
> 
>> I believe the h/w is DECUServer/EncompasServe/name-of-the-weekServe.  I'd
>> assume the h/w and any relevant backups, etc. would be shipped out to any
>> party that would take over its hosting.
> 
> The costs of shipping can be horrendous for computers.
> 
> However, shipping of some peripherals (do do backups etc) might have to
> be involved.
> 
> However, bear in mind that if the current system hoster is extremely
> busy, he may not have the time to "pack" decuserve and ship it.

Again, being only two hours away and owning a truck, I'd be more than 
happy to help with that end of things, if this comes to pass.  I could 
haul it home from there, pack it up, then send it off with a shipper.  I 
buy and sell a lot of weird big stuff on ebay and know my way around 
shipping unwieldy stuff.

If the final destination were anywhere within 8 or 10 hours or so of 
Chicago I'd just haul it there directly and make a day trip of it. 
(Alan, you're out in California, right, so that's not an option here?)

I too had briefly considered offering to host it, but my 768K in the 
outward direction DSL line would seriously disappoint potential users!

Duncan

Duncan Brown wrote:

> I too had briefly considered offering to host it, but my 768K in the 
> outward direction DSL line would seriously disappoint potential users!

My thinking was more along the lines of a temporary backup site being
setup to give time to the main site to get fixed up. As such, 768k
wouldn't be too bad. (Isn't decuserve mostly text anyways ?)

Duncan Brown wrote:

> (Alan, you're out in California, right, so that's not an option here?)

Correct, Los Angeles to be exact.

I like the idea of a backup site, whether "hot" or "cold".
Even a periodic (weekly?) exchange of backup tapes would guard
against the total loss of a site.   Likewise, giving more
than one person access/control of the domain registration
and DNS would be a good thing.

We're supposed to be experts in high-availability systems,
so let's do it!

Alan

In article <006f424f$0$10078$c3e8da3@news.astraweb.com>,
 JF Mezei <jfmezei.spamnot@vaxination.ca> wrote:

> Duncan Brown wrote:
> 
> > I too had briefly considered offering to host it, but my 768K in the 
> > outward direction DSL line would seriously disappoint potential users!
> 
> My thinking was more along the lines of a temporary backup site being
> setup to give time to the main site to get fixed up. As such, 768k
> wouldn't be too bad. (Isn't decuserve mostly text anyways ?)

Yes, DECUServe is mostly text, and 768K is more than enough
to keep a lot of people happy.  However, it's also the home
to some of the Library collections and other files.  I don't
know how often anyone accesses them and downloads them.

Speaking for myself, I think it wouldn't take much bandwidth
to keep DECUServe running.  I'm not in a position to host it
now, but I'll help with whatever I can.

Bart.

"Duncan Brown" <brown_du@eisner.decus.org> wrote in message 
news:RMudnVzKAoECwh7UnZ2dnUVZ_q7inZ2d@speakeasy.net...
> Again, being only two hours away and owning a truck, I'd be more than 
> happy to help with that end of things, if this comes to pass.  I could 
> haul it home from there, pack it up, then send it off with a shipper.  I 
> buy and sell a lot of weird big stuff on ebay and know my way around 
> shipping unwieldy stuff.
>
> If the final destination were anywhere within 8 or 10 hours or so of 
> Chicago I'd just haul it there directly and make a day trip of it. (Alan, 
> you're out in California, right, so that's not an option here?)
>
> I too had briefly considered offering to host it, but my 768K in the 
> outward direction DSL line would seriously disappoint potential users!
>
> Duncan

Thanks, Duncan.  I've passed along your offer. 

Malcolm Dunnett wrote:
> VAXman- @SendSpamHere.ORG wrote:
>> In article <zZKdnbKkAvOX1R_UnZ2dnUVZ_jWWnZ2d@supernews.com>, Alan 
>> Frisbie <Usenet02_Remove@Flying-Disk.com> writes:
>>> Richard B. Gilbert wrote:
>>>
>>>> I don't see anyone else volunteering to install T1 service and host 
>>>> Eisner!
>>> OK, I guess it is time to step up to the plate.   If it becomes
>>> necessary for DECUServe to move, I will give it a new home.
> 
>>
>> I thought about that too Alan.  I have about 20 times your bandwidth but
>> if you want the duties, you're welcome to it and I know you'd do a good
>> job too.
>>
>>
> 
> Yes, thanks Alan. I was also wrestling with whether or not to volunteer 
> to host it (easy to say after someone else has already volunteered I 
> guess).
> 
> I have lots of spare VMS capable hardware here and a 100Mbit pipe, but 
> I'm not sure what sort of bureaucracy I'd have to go through to get my 
> employer (a university) to approve it. I have to admit I've kind of 
> drifted away from Emcompassserve lately anyway(not from VMS though), so 
> I'm not personally suffering from Eisner being down.
> 
> Also, I'm in Canada and hosting it here would go against the current 
> protectionist spirit of the US Congress :-)

Ahhh!!!   But the US Congress has no jurisdiction in Canada.  Just 
ignore them!  Having their egos deflated might be good for them!

JF Mezei wrote:
> VAXman- @SendSpamHere.ORG wrote:
> 
>> I believe the h/w is DECUServer/EncompasServe/name-of-the-weekServe.  I'd
>> assume the h/w and any relevant backups, etc. would be shipped out to any
>> party that would take over its hosting.
> 
> The costs of shipping can be horrendous for computers.
> 
A lot depends on the computer.  I wouldn't want to pay to have a VAX 
11/780 shipped anywhere!!   A MicroVAX 3100 should be pretty cheap to ship.

I got an Alphastation XP1000 shipped to me for $55 US.  It traveled 
about 700 miles!

Duncan Brown wrote:
> Again, being only two hours away and owning a truck

Duncan - could you drop me an e-mail please, thankyou.

On Feb 3, 2:09=A0am, Graham Burley <burley.li...@btinternet.com> wrote:
> Duncan Brown wrote:
> > Again, being only two hours away and owning a truck
>
> Duncan - could you drop me an e-mail please, thankyou.

This is all getting to be somewhat of a joke. While I can understand
and appreciate the effort that Stephen has put in, in the past, and
the fact that he "has a business to run" and "is entering into
politics", it should also be noted that people have offered

a) technical advise on SWXCRs
b) Spare SWXCR's
c) Spare drives
d) Spare Systems
e) Hosting Services/Bandwidth/Colo
f) Hands-on
g)  Shipping to/from alternate locations...

We have seen nothing in nearly two months now... Not even a status
message. Just how long would it take Mr Arnold to pen a "This is what
I have done, this is what I need to do, this is what I need, and this
is when I plan to do it" note...

If he is that distracted by his consulting and political activities,
then he really should hand it all off to someone else.. As has been
mentioned elsewhere, this reflects poorly on the "quality" of the VMS
we know and love..

ANd yes, before anyone grumps at me for not calling him - several
others appear to have, and I am half way around the world...

Mister Q wrote:

> This is all getting to be somewhat of a joke. While I can understand
> and appreciate the effort that Stephen has put in, in the past, and
> the fact that he "has a business to run" and "is entering into
> politics", it should also be noted that people have offered

This is the nature of volunteer work. There are times when work/family
issues tend to overwhelm someone who can't spend ANY time on volunteer
stuff for some period of time. You can't fault a VOLUNTEER for that.



It would help to know how much disk space is needed to host eisner, and
how much monthly data usage is involved for the internet line.


Enough of us have sufficient hardware at home to run this. We can make
all the offers in the world, but there needs to be someone in charge to
decide "OK, we are moving it from X to Y, give X  Y's address and ask X
to send backups to Y.

JF Mezei wrote:
> Mister Q wrote:
> 
>> This is all getting to be somewhat of a joke. While I can understand
>> and appreciate the effort that Stephen has put in, in the past, and
>> the fact that he "has a business to run" and "is entering into
>> politics", it should also be noted that people have offered
> 
> This is the nature of volunteer work. There are times when work/family
> issues tend to overwhelm someone who can't spend ANY time on volunteer
> stuff for some period of time. You can't fault a VOLUNTEER for that.
> 
> 
> 
> It would help to know how much disk space is needed to host eisner, and
> how much monthly data usage is involved for the internet line.
> 
> 
> Enough of us have sufficient hardware at home to run this. We can make
> all the offers in the world, but there needs to be someone in charge to
> decide "OK, we are moving it from X to Y, give X  Y's address and ask X
> to send backups to Y.
> 
> 
> 
This assumes that X, who has evidently abandoned his post, is both able 
and willing to make backups, and to ship the tapes to Y!  It also 
assumes that Y will be able to read the tapes when they get there.  All 
DLT drives are not created equal and tapes are not necessarily 
interchangeable.

There are enterprises that will, for money, host your system, provide 
internet access, etc.  For more money, some will mount tapes for 
backups, send tapes off site, etc, etc.  Does anybody want to spring for it?

Richard B. Gilbert wrote:

> This assumes that X, who has evidently abandoned his post, is both able 
> and willing to make backups, and to ship the tapes to Y!

Since the disks are apparently gone, one would assume that whatever
existing backup will have to be used as basis for the new system.

Is there an expectation that disks in that SWXCR thing-a-ma-jig are
rescuable ?

Remember that if the owner's goal is to ressucitate this system, one has
to understand that debugging disk array problems is not as easy as
figuring out next week's winning lotery numbers. One also has to
consider the possibility that the broken array controller may have
writen bad info and is making recovery extremely difficult.

When the LDdriver zapped one of a list of MOUNT/BIND disks for me, it
took me a lot of hours to figure out what had happened, then build
procedures to inventory the disk, then test each individual files whith
whatever tool I had for its integrity (zip for zip files, ana/rms for
indexed files, type for text files etc etc). Then, from that list, I
then had to rebuild individual files from whatever backup I had, some of
which couldn't be rebuilt because backup was too old.


> It also 
> assumes that Y will be able to read the tapes when they get there.  All 
> DLT drives are not created equal and tapes are not necessarily 
> interchangeable.

Then the volunteer will also have to ship the DLT drive if that is the case.

But right now, this is all speculation. Except for information that this
is a SWXCR based disk array we don't have any additioanl information on
that system.

JF Mezei wrote:
> Richard B. Gilbert wrote:
> 
>> This assumes that X, who has evidently abandoned his post, is both able 
>> and willing to make backups, and to ship the tapes to Y!
> 
> Since the disks are apparently gone, one would assume that whatever
> existing backup will have to be used as basis for the new system.
> 
> Is there an expectation that disks in that SWXCR thing-a-ma-jig are
> rescuable ?

Assuming that the SWXCR was configured for RAID-5, any set of N-1 drives 
in working condition should be sufficient to recover all the data.

We seem to be working with a paucity of reliable information here.

What I gather from reading c.o.v. is:
a. Eisner is down
b. It is in the custody of Steve (somebody) not known to me.
c. Steve is too busy, with things important to him, to deal with the 
problem.
>
I'd suggest replacing any SWXCR with an HSZ or HSG class controller.
I never had a problem with mine but I understand that they have a poor 
reputation!

> Remember that if the owner's goal is to ressucitate this system, one has
> to understand that debugging disk array problems is not as easy as
> figuring out next week's winning lotery numbers. One also has to
> consider the possibility that the broken array controller may have
> writen bad info and is making recovery extremely difficult.
> 
> When the LDdriver zapped one of a list of MOUNT/BIND disks for me, it
> took me a lot of hours to figure out what had happened, then build
> procedures to inventory the disk, then test each individual files whith
> whatever tool I had for its integrity (zip for zip files, ana/rms for
> indexed files, type for text files etc etc). Then, from that list, I
> then had to rebuild individual files from whatever backup I had, some of
> which couldn't be rebuilt because backup was too old.
> 
A volume set created with MOUNT /BIND is in NO WAY equivalent to a RAID 
array as I understand the concept!  The failure of ANY disk in a volume 
set is sufficient to bring down the whole set since any piece of any 
file can be stored on any disk in the set.  The weaknesses of a volume 
set are well known and anyone who puts important data on such a volume 
set deserves anything that happens to that volume set.

In the course of a twenty year career as a VMS system manager I created 
ONE such volume set, being fully aware of its limitations! I used it to 
store Oracle "redo" files which had a useful life of, at most, one day, 
since the database was backed up daily.  If I had had a larger disk 
available, I would have used it in preference to creating that volume set.

<snip>

Richard B. Gilbert wrote:

> Assuming that the SWXCR was configured for RAID-5, any set of N-1 drives 
> in working condition should be sufficient to recover all the data.

Would it be correct to state that in order to recover the data 	, you
would need an SWXCR of the same revision and exact same config to
rebuild the set ?

I realise that MOUNT/BIND is a very primitive way to glue disks
together, but the end result becomes the same: if the software ends up
zapping data in the wrong place, recovery is very hard.

If the problem is the SWXCR itself, then can you ship the hard drives
elsewhere and expect anyone to be able to rebuild the set ?


> We seem to be working with a paucity of reliable information here.

This has *never* stopped c.o.v. from speculating... The paucity of
information about VMS since the Palmer days has trained us to be
extremely good at speculation :-( :-( :-(


> I'd suggest replacing any SWXCR with an HSZ or HSG class controller.

But will those new controllers be able to recover the data ?

In article <0074ae93$0$10071$c3e8da3@news.astraweb.com>, JF Mezei <jfmezei.spamnot@vaxination.ca> writes:
>Richard B. Gilbert wrote:
>
>> Assuming that the SWXCR was configured for RAID-5, any set of N-1 drives 
>> in working condition should be sufficient to recover all the data.
>
>Would it be correct to state that in order to recover the data 	, you
>would need an SWXCR of the same revision and exact same config to
>rebuild the set ?

Not exactly.  The revision isn't as vital - a later rev could work.

But with the configuration -- it depends on how the disks were built.  If
memory serves, when you set up  your RAID  sets in the SWXCR you can sacrifice
a little bit of disk space to save the configuration on the disk itself. This
is portable to other SWXCRs. If you didn't choose the portable configuration,
then you're kind of screwed, because if you tell the SWXCR to build a disk set
it'll trash whatever's on  the disk already.

On the other hand, if Eisner was just using SWXCR to serve JBODs, then I think
those disks should be portable to other controllers.  

>
>I realise that MOUNT/BIND is a very primitive way to glue disks
>together, but the end result becomes the same: if the software ends up
>zapping data in the wrong place, recovery is very hard.
>
>If the problem is the SWXCR itself, then can you ship the hard drives
>elsewhere and expect anyone to be able to rebuild the set ?
>

Depends on whether they were
 - stripesets not configured as portable (then I think your only hope is the
   EEPROM from the dead SWXCR)
 - stripesets configured as portable (then yes, to another SWXCR)
 - JBOD (yes, any controller that can drive the disk)
 - mirrorsets (Raid 0+1, then yes)

>
>> We seem to be working with a paucity of reliable information here.
>
>This has *never* stopped c.o.v. from speculating... The paucity of
>information about VMS since the Palmer days has trained us to be
>extremely good at speculation :-( :-( :-(
>
>
>> I'd suggest replacing any SWXCR with an HSZ or HSG class controller.
>
>But will those new controllers be able to recover the data ?

If you've got an HSZ or HSG class controller, you can use bigger drives and
bigger RAIDsets than the SWXCR can do.  But what you want is to get the data
off the drives - if it's JBOD or mirrorset, should be no problem with HSZ; if
anything else you need SWXCR - and then move it to the less ancient and weird
controller which can manage higher capacity disks.

-- Alan

Alan Winston - SSRL Central Computing wrote:
> In article <0074ae93$0$10071$c3e8da3@news.astraweb.com>, JF Mezei <jfmezei.spamnot@vaxination.ca> writes:
>> Richard B. Gilbert wrote:
>>
>>> Assuming that the SWXCR was configured for RAID-5, any set of N-1 drives 
>>> in working condition should be sufficient to recover all the data.
>> Would it be correct to state that in order to recover the data 	, you
>> would need an SWXCR of the same revision and exact same config to
>> rebuild the set ?
> 
> Not exactly.  The revision isn't as vital - a later rev could work.
> 
> But with the configuration -- it depends on how the disks were built.  If
> memory serves, when you set up  your RAID  sets in the SWXCR you can sacrifice
> a little bit of disk space to save the configuration on the disk itself. This
> is portable to other SWXCRs. If you didn't choose the portable configuration,
> then you're kind of screwed, because if you tell the SWXCR to build a disk set
> it'll trash whatever's on  the disk already.
> 
> On the other hand, if Eisner was just using SWXCR to serve JBODs, then I think
> those disks should be portable to other controllers.  
> 
>> I realise that MOUNT/BIND is a very primitive way to glue disks
>> together, but the end result becomes the same: if the software ends up
>> zapping data in the wrong place, recovery is very hard.
>>
>> If the problem is the SWXCR itself, then can you ship the hard drives
>> elsewhere and expect anyone to be able to rebuild the set ?
>>
> 
> Depends on whether they were
>  - stripesets not configured as portable (then I think your only hope is the
>    EEPROM from the dead SWXCR)
>  - stripesets configured as portable (then yes, to another SWXCR)
>  - JBOD (yes, any controller that can drive the disk)
>  - mirrorsets (Raid 0+1, then yes)
> 
>>> We seem to be working with a paucity of reliable information here.
>> This has *never* stopped c.o.v. from speculating... The paucity of
>> information about VMS since the Palmer days has trained us to be
>> extremely good at speculation :-( :-( :-(
>>
>>
>>> I'd suggest replacing any SWXCR with an HSZ or HSG class controller.
>> But will those new controllers be able to recover the data ?
> 
> If you've got an HSZ or HSG class controller, you can use bigger drives and
> bigger RAIDsets than the SWXCR can do.  But what you want is to get the data
> off the drives - if it's JBOD or mirrorset, should be no problem with HSZ; if
> anything else you need SWXCR - and then move it to the less ancient and weird
> controller which can manage higher capacity disks.
> 
> -- Alan

It has been about five years since I last did anything with a SWXCR but 
ISTR that you could save the configuration on a floppy disk and that 
this floppy would allow you to replace SWXCRs by loading the new with a 
configuration identical to the old.  AIRC I never actually did it.

If I had a choice I would much prefer an HSZ or HSG controller but ISTR 
that I bought SWXCRs on e-Bay for $35 each ca. 2003-2004 and that 
HSZ/HSG class hardware cost somewhere between 50 and 500 times that!

Richard B. Gilbert wrote:
> It has been about five years since I last did anything with a SWXCR but 
> ISTR that you could save the configuration on a floppy disk and that 
> this floppy would allow you to replace SWXCRs by loading the new with a 
> configuration identical to the old.  AIRC I never actually did it.
>

That is correct, you could save the config to floppy which would save 
lots of time if replacing the controller.  Of course, that depends on 
the system manager actively saving the latest config to floppy... and 
from all I've heard I wouldn't bet on that being the case here.

If I had this system in front of me, I would expend a fair amount of 
effort trying to get the existing SWXCR and disks running again. 
Failing that, if it was possible to hand-transcribe the current config 
from the management program (or better yet, save it to floppy) I would 
try replacing the SWXCR and seeing if I could get it running.  If the 
SXWCR is completely dead/corrupt and the config is not discernable then 
I would simply rebuild the system onto newer/less insane storage from 
whatever the latest backups are, and be satisfied with that.

Even if the system is completely recoverable using the existing SWXCR 
and disks, I would migrate off of that setup ASAP.

Duncan

"Duncan Brown" <brown_du@eisner.decus.org> wrote in message 
news:CqSdnSG6aeYRNhTUnZ2dnUVZ_juWnZ2d@speakeasy.net...
> Richard B. Gilbert wrote:
>> It has been about five years since I last did anything with a SWXCR but 
>> ISTR that you could save the configuration on a floppy disk and that this 
>> floppy would allow you to replace SWXCRs by loading the new with a 
>> configuration identical to the old.  AIRC I never actually did it.
>>
>
> That is correct, you could save the config to floppy which would save lots 
> of time if replacing the controller.  Of course, that depends on the 
> system manager actively saving the latest config to floppy... and from all 
> I've heard I wouldn't bet on that being the case here.
>
> If I had this system in front of me, I would expend a fair amount of 
> effort trying to get the existing SWXCR and disks running again. Failing 
> that, if it was possible to hand-transcribe the current config from the 
> management program (or better yet, save it to floppy) I would try 
> replacing the SWXCR and seeing if I could get it running.  If the SXWCR is 
> completely dead/corrupt and the config is not discernable then I would 
> simply rebuild the system onto newer/less insane storage from whatever the 
> latest backups are, and be satisfied with that.
>
> Even if the system is completely recoverable using the existing SWXCR and 
> disks, I would migrate off of that setup ASAP.
>
> Duncan

Absolutely agree on all points.

    Dale 

Duncan Brown wrote:
> Richard B. Gilbert wrote:
>> It has been about five years since I last did anything with a SWXCR 
>> but ISTR that you could save the configuration on a floppy disk and 
>> that this floppy would allow you to replace SWXCRs by loading the new 
>> with a configuration identical to the old.  AIRC I never actually did it.
>>
> 
> That is correct, you could save the config to floppy which would save 
> lots of time if replacing the controller.  Of course, that depends on 
> the system manager actively saving the latest config to floppy... and 
> from all I've heard I wouldn't bet on that being the case here.
> 
> If I had this system in front of me, I would expend a fair amount of 
> effort trying to get the existing SWXCR and disks running again. Failing 
> that, if it was possible to hand-transcribe the current config from the 
> management program (or better yet, save it to floppy) I would try 
> replacing the SWXCR and seeing if I could get it running.  If the SXWCR 
> is completely dead/corrupt and the config is not discernable then I 
> would simply rebuild the system onto newer/less insane storage from 
> whatever the latest backups are, and be satisfied with that.
> 
> Even if the system is completely recoverable using the existing SWXCR 
> and disks, I would migrate off of that setup ASAP.
> 

That takes money!  The SWXCR's chief (only?) virtue is cheap RAID!

Of course you would prefer HSZ or HSG technology but that stuff costs 
money; a great deal of money the last time I looked.  Of course it has 
been a while since I last priced the stuff. . . .

In article <3cWdnY4c_JcRWBTUnZ2dnUVZ_gSdnZ2d@giganews.com>,
	"Richard B. Gilbert" <rgilbert88@comcast.net> writes:
> Duncan Brown wrote:
>> Richard B. Gilbert wrote:
>>> It has been about five years since I last did anything with a SWXCR 
>>> but ISTR that you could save the configuration on a floppy disk and 
>>> that this floppy would allow you to replace SWXCRs by loading the new 
>>> with a configuration identical to the old.  AIRC I never actually did it.
>>>
>> 
>> That is correct, you could save the config to floppy which would save 
>> lots of time if replacing the controller.  Of course, that depends on 
>> the system manager actively saving the latest config to floppy... and 
>> from all I've heard I wouldn't bet on that being the case here.
>> 
>> If I had this system in front of me, I would expend a fair amount of 
>> effort trying to get the existing SWXCR and disks running again. Failing 
>> that, if it was possible to hand-transcribe the current config from the 
>> management program (or better yet, save it to floppy) I would try 
>> replacing the SWXCR and seeing if I could get it running.  If the SXWCR 
>> is completely dead/corrupt and the config is not discernable then I 
>> would simply rebuild the system onto newer/less insane storage from 
>> whatever the latest backups are, and be satisfied with that.
>> 
>> Even if the system is completely recoverable using the existing SWXCR 
>> and disks, I would migrate off of that setup ASAP.
>> 
> 
> That takes money!  The SWXCR's chief (only?) virtue is cheap RAID!

And, what good is RAID that leaves the system down for months at a time?

> 
> Of course you would prefer HSZ or HSG technology but that stuff costs 
> money; a great deal of money the last time I looked.  Of course it has 
> been a while since I last priced the stuff. . . .

I think mine is HSJ based but I have a StorageWorks Cabinet that cost
the same as my many VAXen, Alpha's and PDP-11's (Hint: nothing unless
you count the cost of transporting it which was usually only gas).

bill

-- 
Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolves
billg999@cs.scranton.edu |  and a sheep voting on what's for dinner.
University of Scranton   |
Scranton, Pennsylvania   |         #include <std.disclaimer.h>   

Richard B. Gilbert wrote:
> Duncan Brown wrote:
>> Richard B. Gilbert wrote:
>>> It has been about five years since I last did anything with a SWXCR 
>>> but ISTR that you could save the configuration on a floppy disk and 
>>> that this floppy would allow you to replace SWXCRs by loading the new 
>>> with a configuration identical to the old.  AIRC I never actually did 
>>> it.
>>>
>>
>> That is correct, you could save the config to floppy which would save 
>> lots of time if replacing the controller.  Of course, that depends on 
>> the system manager actively saving the latest config to floppy... and 
>> from all I've heard I wouldn't bet on that being the case here.
>>
>> If I had this system in front of me, I would expend a fair amount of 
>> effort trying to get the existing SWXCR and disks running again. 
>> Failing that, if it was possible to hand-transcribe the current config 
>> from the management program (or better yet, save it to floppy) I would 
>> try replacing the SWXCR and seeing if I could get it running.  If the 
>> SXWCR is completely dead/corrupt and the config is not discernable 
>> then I would simply rebuild the system onto newer/less insane storage 
>> from whatever the latest backups are, and be satisfied with that.
>>
>> Even if the system is completely recoverable using the existing SWXCR 
>> and disks, I would migrate off of that setup ASAP.
>>
> 
> That takes money!  The SWXCR's chief (only?) virtue is cheap RAID!
> 
> Of course you would prefer HSZ or HSG technology but that stuff costs 
> money; a great deal of money the last time I looked.  Of course it has 
> been a while since I last priced the stuff. . . .
> 
> 

Well, I had built myself a completely redundant completely outfitted 
HSJ50 setup out of used parts for nearly nothing, and sold it again a 
couple of years ago for nearly nothing, so that could be repeated I'm 
sure.  But at least one of the offers for alternate hosting site seen 
here included existing storage.  And as I said a couple of weeks ago, I 
think at this point we'd all be better served by single disks and a good 
backup strategy, than by a SWXCR.  I mean really...for something like 
Eisner (as opposed to a business-critical system) you might suffer, 
what, a dead disk every two or three years, and a couple of hours of 
hassle and $100 to recover from it?

Duncan

Bill Gunshannon wrote:
> In article <3cWdnY4c_JcRWBTUnZ2dnUVZ_gSdnZ2d@giganews.com>,
> 	"Richard B. Gilbert" <rgilbert88@comcast.net> writes:
>> Duncan Brown wrote:
>>> Richard B. Gilbert wrote:
>>>> It has been about five years since I last did anything with a SWXCR 
>>>> but ISTR that you could save the configuration on a floppy disk and 
>>>> that this floppy would allow you to replace SWXCRs by loading the new 
>>>> with a configuration identical to the old.  AIRC I never actually did it.
>>>>
>>> That is correct, you could save the config to floppy which would save 
>>> lots of time if replacing the controller.  Of course, that depends on 
>>> the system manager actively saving the latest config to floppy... and 
>>> from all I've heard I wouldn't bet on that being the case here.
>>>
>>> If I had this system in front of me, I would expend a fair amount of 
>>> effort trying to get the existing SWXCR and disks running again. Failing 
>>> that, if it was possible to hand-transcribe the current config from the 
>>> management program (or better yet, save it to floppy) I would try 
>>> replacing the SWXCR and seeing if I could get it running.  If the SXWCR 
>>> is completely dead/corrupt and the config is not discernable then I 
>>> would simply rebuild the system onto newer/less insane storage from 
>>> whatever the latest backups are, and be satisfied with that.
>>>
>>> Even if the system is completely recoverable using the existing SWXCR 
>>> and disks, I would migrate off of that setup ASAP.
>>>
>> That takes money!  The SWXCR's chief (only?) virtue is cheap RAID!
> 
> And, what good is RAID that leaves the system down for months at a time?
> 

Not much but I don't think that RAID is entirely at fault.  It appears 
from discussion here that the custodian of Eisner has abandoned his post.

If anyone had been "minding the store" the problem almost certainly 
could have been fixed three or four weeks ago!  Either RAID-1 or RAID-5
requires the failure of at least TWO disks to bring the RAID set down!
Failure of the controller, of course, will also do it.  The SWXCR is 
hardly the most reliable technology available these days!

ISTR that a failed RZ series disks gives warning by lighting an amber 
LED on the front panel; a single glance tells you "dead disk"!

But speculation is not going to get anything fixed.  Neither is posting 
to c.o.v/info-vax.

1.  A new custodian for Eisner is needed.
2.  Someone must reclaim the mortal remains of Eisner from its current 
custodian.
3.  The new custodian must gain possession.
4.  The new custodian must repair what he can and get someone to repair 
anything he cannot!
5.  The new custodian must either have or must install the necessary 
communications bandwidth!
6.  Plug it in, restore, and go!

Given the history so far, I'm not going to hold my breath!

<snip>

On 2009-02-03, Richard B. Gilbert <rgilbert88@comcast.net> wrote:
> This assumes that X, who has evidently abandoned his post, is both able 
> and willing to make backups, and to ship the tapes to Y!  It also 
> assumes that Y will be able to read the tapes when they get there.  All 
> DLT drives are not created equal and tapes are not necessarily 
> interchangeable.
>

Backups were done on a regular basis (but I don't know the schedule), so
if all else fails, it should just be a matter of restoring from the last
backup made before the system failure.

As for the question you raised in another message regarding the SWXCR
(which, BTW, I dislike as much as you apparently do :-) ), there was talk
on the mailing list about doing down the host based volume shadowing
route. (Eisner is run using the hobbyist licenses and I understand volume
shadowing is available as part of the hobbyist program.)

Simon.

-- 
Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP
Microsoft: Bringing you 1980's technology to a 21st century world
(Note: Email address not currently working)

Simon Clubley wrote:
> On 2009-02-03, Richard B. Gilbert <rgilbert88@comcast.net> wrote:
>> This assumes that X, who has evidently abandoned his post, is both able 
>> and willing to make backups, and to ship the tapes to Y!  It also 
>> assumes that Y will be able to read the tapes when they get there.  All 
>> DLT drives are not created equal and tapes are not necessarily 
>> interchangeable.
>>
> 
> Backups were done on a regular basis (but I don't know the schedule), so
> if all else fails, it should just be a matter of restoring from the last
> backup made before the system failure.
> 
> As for the question you raised in another message regarding the SWXCR
> (which, BTW, I dislike as much as you apparently do :-) ), there was talk
> on the mailing list about doing down the host based volume shadowing
> route. (Eisner is run using the hobbyist licenses and I understand volume
> shadowing is available as part of the hobbyist program.)

Having some experience with host based volume shadowing, I think I may 
hate host based volume shadowing as much as some people hate the SWXCR! 
  My experience with HBVS is about ten years out of date now and things 
may have improved.  Problems with HBVS were seemingly endless!  My 
experience is about ten years old now; we were forced into HBVS when
DEC dropped support for controller based (HSC-50/70/90) volume shadowing.

Fortunately for the sanity of all concerned, the HSC controllers and the 
RA series disks were ALMOST history by 1998!  Most of the production 
systems were using HSZ or HSJ controllers and RZ series disks.

In article <S-ydne1UePCwdBTUnZ2dnUVZ_qjinZ2d@giganews.com>, Richard B. Gilbert 
wrote:
[...]
>
>Having some experience with host based volume shadowing, I think I may 
>hate host based volume shadowing as much as some people hate the SWXCR! 
>  My experience with HBVS is about ten years out of date now and things 
>may have improved.  Problems with HBVS were seemingly endless!  My 
>experience is about ten years old now; we were forced into HBVS when
>DEC dropped support for controller based (HSC-50/70/90) volume shadowing.

FWIW, I've used HBVS for a number of years, on both mission-critical systems,
and on my hobbyist system at home.  I've never experienced a failure mode that
was impossible to recover from.  99% of the time, it was/is a simple matter of
replacing the failed component, and allowing shadowing to "kick in".  HBVS has
definitely improved since the last century.
[...]

Bradford Hamilton wrote:
> In article <S-ydne1UePCwdBTUnZ2dnUVZ_qjinZ2d@giganews.com>, Richard B. Gilbert 
> wrote:
> [...]
>> Having some experience with host based volume shadowing, I think I may 
>> hate host based volume shadowing as much as some people hate the SWXCR! 
>>  My experience with HBVS is about ten years out of date now and things 
>> may have improved.  Problems with HBVS were seemingly endless!  My 
>> experience is about ten years old now; we were forced into HBVS when
>> DEC dropped support for controller based (HSC-50/70/90) volume shadowing.
> 
> FWIW, I've used HBVS for a number of years, on both mission-critical systems,
> and on my hobbyist system at home.  I've never experienced a failure mode that
> was impossible to recover from.  99% of the time, it was/is a simple matter of
> replacing the failed component, and allowing shadowing to "kick in".  HBVS has
> definitely improved since the last century.
> [...]
> 

My last experience was about ten years ago.  I hope things are different 
now!

Duncan Brown <brown_du@eisner.decus.org> wrote:
> JF Mezei wrote:
>> VAXman- @SendSpamHere.ORG wrote:
>> 
>>> I believe the h/w is DECUServer/EncompasServe/name-of-the-weekServe.  I'd
>>> assume the h/w and any relevant backups, etc. would be shipped out to any
>>> party that would take over its hosting.
>> 
>> The costs of shipping can be horrendous for computers.

[...]
 
> Again, being only two hours away and owning a truck, I'd be more than 
> happy to help with that end of things, if this comes to pass.  I could 
> haul it home from there, pack it up, then send it off with a shipper.  I 
> buy and sell a lot of weird big stuff on ebay and know my way around 
> shipping unwieldy stuff.

I'm willing to help as well, and I'm just slightly over three hours
away.  Maybe this weekend?

> If the final destination were anywhere within 8 or 10 hours or so of 
> Chicago I'd just haul it there directly and make a day trip of it. 
> (Alan, you're out in California, right, so that's not an option here?)

California is a bit far to deliver to - shipping a DS20 would be
a far more reasonable option.  I doubt we are so strapped that we
cannot come up with a few bucks for shipping.

> I too had briefly considered offering to host it, but my 768K in the 
> outward direction DSL line would seriously disappoint potential users!

Same situation here.  I might be willing to host it for a year or so,
but eventually I could find myself in a similar situation to Stephen's,
with other burdens keeping me from being able to spend a sufficient
amount of time on it to stay ahead of whatever problems might arise.

Since I work with two other VMS system managers, I could perhaps rely
on them when I was out of town to keep the system running, and that
could be a plus.  Lots of multidatacenter clustering experience here
(HS-series controllers and HBVS).

Now if I could just get the boss to part with a couple of the HSG80
controllers we'll be decommissioning next month...

-- 
George Cornelius                          cornelius AT mayo DOT edu

> Duncan

> Now if I could just get the boss to part with a couple of the HSG80
> controllers we'll be decommissioning next month...

Nice, but overkill for this application, I think.  We need to consider
power consumption as well.

As for the comments about host-based shadowing in other 
articles: I used it for the last few years while I was in
OpenVMS engineering, and it worked really well.  I had
two-disk shadow pairs for the department server, and
while we had a few disks go bad, it always recovered
pretty much automatically.

Bart.

On Feb 5, 4:37=A0am, ZL <bz...@copper.net> wrote:
[...]
>
> As for the comments about host-based shadowing in other
> articles: I used it for the last few years while I was in
> OpenVMS engineering, and it worked really well. =A0I had
> two-disk shadow pairs for the department server, and
> while we had a few disks go bad, it always recovered
> pretty much automatically.

Let me just add, having used HBVS for 15+ years, that
the last major changes, which were made in the VS 7.3-2
time-frame and which added HBMM, have *greatly*
improved the performance during shadow copies and
shadow merges.

I managed a mixed VAX-Alpha cluster on VMS 6/6.2-1H3.
A crashed system (or even a normal shutdown in some
cases!) would result in the whole disk farm going into *full*
merge (we had HSC-50's mainly).  The system preformance
would turn to molasses for hours afterward!

Late in the VMS 6.x time frame, a shadowing patch fixed
some of the mixed cluster issues so that HSC-assisted
mini-merges would work again (whew!).

But a little later, we had ES40's on HSG-80s and there
was *no* mini-merge.  Again, it was painful to have a
system crash (I think it was like 48-72 hours to complete
the full merge of that much larger disk farm).  We lived
with that for almost five years until VMS Engineering
finally got HBMM out the door.  With HBMM, there were
other improvements to the merge and copy algorithms
that both did a better job of not impacting system
performance *and* sped up the merges.

So I'd say wrt to HBVS, "go for it".  :-)

    -Ken

In article <slrngokcqs.2v1u.hamilton@rabbit.turquoisewitch.com>,
hamilton@news.individual.net (Bradford Hamilton) writes: 

> FWIW, I've used HBVS for a number of years, on both mission-critical systems,
> and on my hobbyist system at home.  I've never experienced a failure mode that
> was impossible to recover from.  99% of the time, it was/is a simple matter of
> replacing the failed component, and allowing shadowing to "kick in".  HBVS has
> definitely improved since the last century.
> [...]

I couldn't agree more.  Especially with the new MINICOPY and MINIMERGE 
functionality.

"Richard B. Gilbert" <rgilbert88@comcast.net> writes:

>Bradford Hamilton wrote:
>> In article <S-ydne1UePCwdBTUnZ2dnUVZ_qjinZ2d@giganews.com>, Richard B. Gilbert 
>> wrote:
>> [...]
>>> Having some experience with host based volume shadowing, I think I may 
>>> hate host based volume shadowing as much as some people hate the SWXCR! 
>>>  My experience with HBVS is about ten years out of date now and things 
>>> may have improved.  Problems with HBVS were seemingly endless!  My 
>>> experience is about ten years old now; we were forced into HBVS when
>>> DEC dropped support for controller based (HSC-50/70/90) volume shadowing.
>> 
>> FWIW, I've used HBVS for a number of years, on both mission-critical systems,
>> and on my hobbyist system at home.  I've never experienced a failure mode that
>> was impossible to recover from.  99% of the time, it was/is a simple matter of
>> replacing the failed component, and allowing shadowing to "kick in".  HBVS has
>> definitely improved since the last century.
>> [...]
>> 

>My last experience was about ten years ago.  I hope things are different 
>now!

Depending on exactly when "about" was, they are different now.  It was
determined that there were some fundamental problems with shadowing and
it was determined it was better to redesign many major components, and
the code was essentially rewritten.  I was part of that effort.  Since
I left, they've made further improvements, write bitmaps and minicopy etc.

ZL <bzl52@copper.net> wrote:
> 
>> Now if I could just get the boss to part with a couple of the HSG80
>> controllers we'll be decommissioning next month...
> 
> Nice, but overkill for this application, I think.  We need to consider
> power consumption as well.

Overkill, of course.  But we need to use what's available
to us (and we've tried underkill!)

Actually, the specs say 23 watts for the controller and 2.5
for the cache (9 when charging the battery).  This doesn't
count the rest of the box, but I would guess that it is
less power than the drives for, say, a seven drive raidset.

0.030 kw x 24 hours/day x $0.10/kwh = $0.072/day
(higher in California, assuming Alan doesn't have a
wind farm on the Altamont pass), so a single controller,
at least, is affordable.

There's more infrastructure needed, though, including
fibre cards, and, unless FC-AL is an option, switches.

> As for the comments about host-based shadowing in other 
> articles: I used it for the last few years while I was in
> OpenVMS engineering, and it worked really well.

Should work fine.

--
George Cornelius              cornelius AT eisner DOT decus DOT org
                              cornelius AT mayo DOT edu

> Bart.

In article <0077e80f$0$11573$c3e8da3@news.astraweb.com>,
 JF Mezei <jfmezei.spamnot@vaxination.ca> wrote:

> Mister Q wrote:
> 
> > This is all getting to be somewhat of a joke. While I can understand
> > and appreciate the effort that Stephen has put in, in the past, and
> > the fact that he "has a business to run" and "is entering into
> > politics", it should also be noted that people have offered
> 
> This is the nature of volunteer work. There are times when work/family
> issues tend to overwhelm someone who can't spend ANY time on volunteer
> stuff for some period of time. You can't fault a VOLUNTEER for that.

I must disagree. While a volunteer might in some situations back out of 
their volunteer commitment, sometimes on short notice, it is not 
acceptable for a volunteer to just go silent. A volunteer has an 
obligation to provide a minimal response even if it's nothing more than 
"I'm too busy. Come take it over." 

How well a volunteer does his or her job impacts his or her reputation. 
Would you hire someone for a position of responsibility if you knew they 
royally screwed up a volunteer job and their attitude was essentially "I 
was volunteer so tough"? I wouldn't.

I say this as a volunteer leader in two different organizations. I am 
well aware of the impact if I suddenly stopped doing my job and didn't 
answer inquiries. I dare say if I suddenly "went silent" on my volunteer 
positions, the impact is probably larger than if I suddenly stopped 
showing up at my real job.

-- 
Larry Stone
lstone*nospam*19@stonejongleux.com

Michael Moroney wrote:
> "Richard B. Gilbert" <rgilbert88@comcast.net> writes:
> 
>> Bradford Hamilton wrote:
>>> In article <S-ydne1UePCwdBTUnZ2dnUVZ_qjinZ2d@giganews.com>, Richard B. Gilbert 
>>> wrote:
>>> [...]
>>>> Having some experience with host based volume shadowing, I think I may 
>>>> hate host based volume shadowing as much as some people hate the SWXCR! 
>>>>  My experience with HBVS is about ten years out of date now and things 
>>>> may have improved.  Problems with HBVS were seemingly endless!  My 
>>>> experience is about ten years old now; we were forced into HBVS when
>>>> DEC dropped support for controller based (HSC-50/70/90) volume shadowing.
>>> FWIW, I've used HBVS for a number of years, on both mission-critical systems,
>>> and on my hobbyist system at home.  I've never experienced a failure mode that
>>> was impossible to recover from.  99% of the time, it was/is a simple matter of
>>> replacing the failed component, and allowing shadowing to "kick in".  HBVS has
>>> definitely improved since the last century.
>>> [...]
>>>
> 
>> My last experience was about ten years ago.  I hope things are different 
>> now!
> 
> Depending on exactly when "about" was, they are different now.  It was
> determined that there were some fundamental problems with shadowing and
> it was determined it was better to redesign many major components, and
> the code was essentially rewritten.  I was part of that effort.  Since
> I left, they've made further improvements, write bitmaps and minicopy etc.
> 

"about ten years ago" means 1998 and VMS/VAX/Alpha V5.5-2 (I think) 
Somewhere around then we upgraded all our machines to V6.1.  AIRC we got 
disks stuck somewhere between mounted and dismounted and fun stuff like 
that!

I remember that there was pain but I've forgotten exactly where it hurt.

"Richard B. Gilbert" <rgilbert88@comcast.net> writes:

>Michael Moroney wrote:

>>> My last experience was about ten years ago.  I hope things are different 
>>> now!
>> 
>> Depending on exactly when "about" was, they are different now.  It was
>> determined that there were some fundamental problems with shadowing and
>> it was determined it was better to redesign many major components, and
>> the code was essentially rewritten.  I was part of that effort.  Since
>> I left, they've made further improvements, write bitmaps and minicopy etc.
>> 

>"about ten years ago" means 1998 and VMS/VAX/Alpha V5.5-2 (I think) 
>Somewhere around then we upgraded all our machines to V6.1.  AIRC we got 
>disks stuck somewhere between mounted and dismounted and fun stuff like 
>that!

>I remember that there was pain but I've forgotten exactly where it hurt.

That was before the rewrite.

In article <BBtaMJj6$rJD@eisner.encompasserve.org>, koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes:
>
>   Thanks to all involved.

Nearly 4 months but it's back!!!  :D


-- 
VAXman- A Bored Certified VMS Kernel Mode Hacker    VAXman(at)TMESIS(dot)ORG

  http://www.quirkfactory.com/popart/asskey/eqn2.png
  
  "Well my son, life is like a beanstalk, isn't it?"

   Thanks to all involved.