ECP VMSDSK Disk Statistics Dara Record
I am using ECP to collect statistical data for performance tracking and
I am currently researching a performance problem with one of the disk drives.
My suspicion is that the disk, $1$DGA3:, is over-burdened with activity.
In the ECP VMSDSK record there is a field for "Disk utilization".
In question is the "Percentage of time disk was busy from perspective of this
The aforementioned disk is cluster mounted (SAN) and the values for the busy
state are 100% for NODE1, 30% for NODE2, and 55% for NODE3. When combined all
three nodes never exceed 300%.
My ...sar disk data
We have several servers running version 4.3.3 and we're collecting disk
performance data with sar. On some of the servers we have noticed negative
values in the sar output for various fields. Does anyone know what may be
causing the negative values and/or how to correct this situation? TIA.
"Janis K" <firstname.lastname@example.org> wrote in message
> We have several servers running version 4.3.3 and we're collecting disk
> performance data with sar. On some of the servers we have noticed
> values in the ...MONITOR DISK statistics
We use a 2-node distributed OpenVMS (Alpha) cluster. The shared disks
are mirrored, so in case of a single-node failure all data is still
available. While trying to measure the system throughput we got
some (IMHO) strange output from MONITOR DISK. It shows all shared
disks three times on each node:
- one for local access,
- one for remote access via foreign node,
- one for 'remote' access via exactly the same node!
I am relatively new to VMS, so maybe my opinion is naive, but
I really cannot understand why do we get a large bulk of
IO going through MSCP server running on exac...sar output & hard disk layout
The machine is an x86 with SunOS 5.9.
I ran 'sar -d' to monitor disk activity, now I am trying to understand
There are some devices like md18 or md20, but also sd2 and sd2,a and
similar, all of which seem to be magnetic disks or slices of such.
The sar man page instructs me to read 'System Administration Guide:
Basic Administration', but I could not find an explanation for the names.
Being familiar from other UNIX flavours, I looked for /dev/sd2 or
/dev/md18, but can only find /dev/md/[r]dsk, /dev/dsk and entries like
/dev/sd128a, all of which are symbolic links...How to see current hard disk i/o statistics?
Is there a possibility to see the hard disk i/o statistics for each process
(like the CPU usage with top(1))?
Thanks in advance
...sar -d and how to determine which filesystem using the disk devices?
One can see from the sar -d reports that md1 is at 100% used.
But how to determine which filesystem using the disk devices? Thanks.
09:06:04 md0 65 1.8 19 365 32.0 63.0
md1 100 11.4 284 7107 1.4 38.8
md10 52 0.8 10 220 0.0 84.2
md11 100 7.6 169 5260 0.0 45.2
md20 37 0.5 10 145 0.0 45.7
md21 100 5.9 169 5260 0.0 35.0
sd2 100 6.7 ..."sar -b" monitoring disk usage
Does anyone know why "sar -b" returns all 0s?? This is the case for most or
all of the machine on my school network. I need to use this tool to detect
throttling during usage.
There are other options such as using "sar -d" or "iostat", however, I would
rather see "sar -b" run properly.
Ji Bae wrote:
> Hello everyone.
> Does anyone know why "sar -b" returns all 0s?? This is the case for most or
> all of the machine on my school network. I need to use this tool to detect
&...Sar showing large disk activity for 1 second periods
I've been looking at some sar logs on a couple of machines (RH8) and I
keep getting some huge peaks - it looks to me like some sort of bug in
sar, but I don't know if anyone has seen this before or not.
I did sar -p -d at intervals of 1 second and caught it happening when
my await went up to 60M when it is normally about 20. These happen ever
hour or so (but not exactly on the hour, and not tallying up with any
08:31:41 DEV tps rd_sec/s wr_sec/s avgrq-sz avgqu-sz await
08:31:42 cciss/c0d0 3.00 0.00 96.00 32.00 0....disk to disk backup
I've got a disk with plenty of data that needs making a backup. Since
new hard disks aren't very expansive we decided that we are going to buy
another disk (identical ATA disk) and that we will do a backup of data
to it. It seems to me that a good starting point would be making a
mirror of the current disk to the new one, but what next? It doesn't
seem acceptable to mirror the disk every day.. Some kind of differential
method would be appropriate. I know that a lot of tools exists that can
do such a things with tapes, but it looks like they use tar and other