f



High system cpu (vs. user cpu)

One of our 9.2.0.6 databases running on RedHat 4.0 has been using a
lot of 'system' cpu today.

12:48:22 PM  CPU   %user   %nice %system %iowait    %irq   %soft
%idle    intr/s
12:48:24 PM  all   47.07    0.00   47.69    0.00    0.25    3.00
2.00   6767.84
12:48:26 PM  all   44.25    0.00   50.88    0.00    0.25    3.12
1.50   6765.17
12:48:28 PM  all   45.58    0.00   49.94    0.00    0.50    2.74
1.25   7071.00
12:48:30 PM  all   46.88    0.00   47.63    0.00    0.50    3.37
1.62   7268.66

All of our db's use MTS, but what's weird (and unique) about this one,
is that it has 3000+ connections into it, and the dispatchers (not the
servers as on other boxes) that are taking up the 'top' cpu
utilization.

Usually that's an indication of a 'spinning' dispatcher to me (and
usually I would find a session in waiting for 'virtual circuit status'
that I'd have to kill to 'free up' the dispatcher).. but today
something is just different.. can't figure it out.. the box is under
higher load then usual, but not 3-4x (seeing 3-4x higher runq) 
anyone seen this? any clue?

Thanks,
andrey
........
We run Oracle 9.2.0.6 on RH4 AMD
remove NSPAM to email
0
4/3/2006 4:48:47 PM
comp.databases.oracle.server 22978 articles. 1 followers. Post Follow

0 Replies
364 Views

Similar Articles

[PageSpeed] 11

Reply:

Similar Artilces:

How , system cpu and user cpu times calculates
Dear friend : I have doubt How , system cpu and user cpu times calculates Thanks On 27 Feb, 06:17, pavunkumar <pavun....@gmail.com> wrote: > I have doubt ["question" in standard english] =A0How , system cpu and =A0 > user cpu times calculates this varies from operating to operating system so you need to ask your question in a news group specific to your OS (eg. Windows or Linux) ...

Very high CPU on OS side vs very high dbFileSeqRead on Oracle side...
Hi, We have a DB in 8.1.7.1 on Sun/Solaris that endures pretty high reads activity, principally located on a few index files. I have to say that these blocks are constantly updated as well as a few tables are inserted *all the time* as well. The application response time is correct, according to our users. The OS CPU consumption is very high (see below). It only leads sometimes -but WE DON'T KNOW WHY- to some delay between the last data available displayed to the users (this is an ASP app.) and the current time (it should never be > 2 mn., yet we sometimes reach 20 minutes of difference => I mean this INSERT activity is slowed down in some way). Steve Adams' script response_time_breakdown.sql shows us 74% of waits concentrated on 'db file seq. read' event (SEVENTY FOUR percent). I determined that these waits principally happen on 2 or 3 index files. My question: IS THERE A WAY TO RELIEVE A BIT the activity on these index blocks which are constantly read ? Data: ===== Nb of client sessions: # ps -ef|grep SID|grep LOCAL=NO|wc -l 129 Here is the hardware: # /usr/platform/sun4u/sbin/prtdiag|more System Configuration: Sun Microsystems sun4u Sun Fire 280R (2 X UltraSPARC-III+) System clock frequency: 150 MHz Memory size: 4096 Megabytes Doing a 'sar' *always* shows you that kind of stuff: ^^^^^^ ARWP>sar 2 20 SunOS ... 5.8 Generic_108528-20 sun4u 03/04/04 10:08:52 %usr %sys ...

fullstimer: user CPU time vs. system time?
I want to reduce the run time of a few big programs, so I'm using the fullstimer system option. I get data back like: I think I understand the real time measurement (clock-on-the-wall time), but what is the difference between user CPU time and system CPU time? I use a multiprocessor CPU linux server. Is the system CPU time the sum of usage on all different CPUs? Also, I'm reading the SUGI30 paper on randomly selecting observations, and the author uses a term "Wall time" in his benchmarks, but I don't know what that is either. TIA Rusty, I'm going to preface m...

Re: fullstimer: user CPU time vs. system time?
Rusty, If you are on a Windows platform, you can also download our log analysis utility to help you break the metrics out of the log. http://utilities.saviansoftware.com It is free, is GUI based, and has an installshield. Thanks, Alan Savian "Bridging SAS and Microsoft Technologies" -----Original Message----- From: SAS(r) Discussion [mailto:SAS-L@LISTSERV.UGA.EDU] On Behalf Of Rusty Shackleford Sent: Thursday, April 14, 2005 2:47 PM To: SAS-L@LISTSERV.UGA.EDU Subject: fullstimer: user CPU time vs. system time? I want to reduce the run time of a few big programs, so I'm using the fullstimer system option. I get data back like: I think I understand the real time measurement (clock-on-the-wall time), but what is the difference between user CPU time and system CPU time? I use a multiprocessor CPU linux server. Is the system CPU time the sum of usage on all different CPUs? Also, I'm reading the SUGI30 paper on randomly selecting observations, and the author uses a term "Wall time" in his benchmarks, but I don't know what that is either. TIA ...

Oracle.exe=high CPU utilization on database logoff? Yes, logoff!
Hey everyone, I've got a question that I can't seem to figure out; maybe you can help? I've got an Oracle 8i database on a Windows 2000 Server o/s. For a while, when I was done with an SQL Plus session on my workstation, if I did a "disconnect", it took about 9 seconds for me to get the SQL prompt back. I just thought my system was slow... Well, I wish I ran to the server as soon as I noticed that- it's not my workstation. It's the server itself. If I connect to SQL Plus, with all the performance counters running, the server (cpu utilization, and all processes' utlilization) barely blinks an eye. As soon as I disconnect from SQL Plus, the CPU utlization jumps to 70%, falls back to about 50%, then jumps up to about 60%, and as soon as the SQL prompt comes back, CPU utilization goes to zero again. I ran the counters for all the oracle background processes (PMON, LGWR, etc), and they don't move; they're near zero. It's oracle.exe itself that has high CPU % !! What could possibly be happening for there to be such a load for a logoff? The database runs fast otherwise (large queries included), but that logoff is the culprit. Even if I log on to sql plus, and quickly disconnect without doing ANY dml/dll/select ops, the delay is there. I even booted everyone off the system on Sunday night (okay, well, nobody was on but me), and the same thing happens. That was the extent of my real-life tests, by the way- I opened one sqlplu...

Re: fullstimer: user CPU time vs. system time? #4
SAS also has the new ARM macros, which purport to measure and report many metrics. See <http://support.sas.com/rnd/scalability/tools/arm/arm.html>. When I last tried the macros, they were not completely implemented, but that might have changed by now. The process looks more confusing than it needs to be, and it doesn't capture some measures that LOGPARSE does (such as number of observations), but it's built into the system at a lower level. > -----Original Message----- > From: SAS(r) Discussion [mailto:SAS-L@LISTSERV.UGA.EDU] On > Behalf Of Michael Raithel > Sent...

How much memory and CPU does the HP server need for a 4 terabytes Oracle database
I am asked to provide a ball park estimate for how many CPUs and how much memory are needed for a 4 terabytes disk space Oracle database running under HP Unix server? Any helpful information is greatly appreciated. Thanks in advance. Peter vu wrote: > I am asked to provide a ball park estimate for how many CPUs and how > much memory are needed for a 4 terabytes disk space Oracle database > running under HP Unix server? Any helpful information is greatly > appreciated. > > Thanks in advance. If you do mainly "select 1 from DUAL", not much... ( ;) * 0.5 ) The need for memory is predecated on how much data you need to be in memory to make queries fast. If a day-to-day usage pattern will query some quantifiable and colocatable subset of the data, then that's about how much memory you need. If you are going to be constantly querying the whole DBMS contents, you will need as much as you can get, and it'll still be slow. Joe Weinstein at BEA Peter vu wrote: > I am asked to provide a ball park estimate for how many CPUs and how > much memory are needed for a 4 terabytes disk space Oracle database > running under HP Unix server? Any helpful information is greatly > appreciated. > > > Thanks in advance. Errrr ... Eight should do! Actually - Your question is wrong. The disk size of the database is almost irrelevant. You need to get an idea of the workload. Is this a warehouse or transaction system; how many us...

Re: fullstimer: user CPU time vs. system time? #6 1546728
Rusty, I'm going to preface my comments by saying that I don't have any experience looking at these timings on linux, so there may be some OS differences here. System CPU time on MVS reports things like involuntary swap time (when your job is swapped out while waiting on I/O, for instance) and generating the I/O request itself (looks in memory first, to see if it already has what you're looking for--if it's not there, generates and sends the I/O request). For multiprocessors, the system time includes all the overhead of keeping track of all the work performed by the system (splitting and pulling together the parallel threads). Typically, these system times are small, and in general, there's not a lot you can do about them. You can attempt to tweak your bufno and memsize/memleave options, but unless you can identify that your problem is system related (and not code/data structure related), you'll probably do better to focus on the other, more tunable things first. Having said that, if you see that the time is significant, you might look to see if there are knobs in linux that you can turn, to try and isolate/boost the priority of your job. In nearly every efficiency study I've ever done, when running against large datasets, you're nearly always bound by I/O. Use some of Michael Raithel's suggestions on efficiency--use only the variables you need, only the records you need, don't do any transformations unless you have to, try to u...

High CPU load but CPU idle
We have HP-UX 11.0 on N class machines. Apache does not run correctly on this machine and HP has no clue why the OS acts this way. CPU 0 has a very high load but is idle. Cpu states: CPU LOAD USER NICE SYS IDLE BLOCK SWAIT INTR SSYS 0 41.95 0.0% 0.0% 0.0% 100.0% 0.0% 0.0% 0.0% 0.0% Anyone have any ideas what's going on? Any time this happens our only option is to reboot. In article <1171384883.637556.310370@v33g2000cwv.googlegroups.com>, tony & jean <vecchiet@gmail.com> wrote: >We have HP-UX 11.0 on N class machines. >Apache do...

oracle database -- multiple users vs single user
Hi Oracle/JDBC guru, What is the difference for java/database project by using multiple users vs. single user? Our applications using oracle 9i database and the DBA created multiple users to manage different sets of tables. But all users have to grant insert, update, delete privileges to each other. There is an overhead on DBA task obviously. Does the multiple users mechanism has any benefits on JDBC part of the application? For example, a better connections? (There is a connection pool) yu2791 wrote: > Hi Oracle/JDBC guru, > > What is the difference for java/database project by using multiple > users vs. single user? > > Our applications using oracle 9i database and the DBA created multiple > users to manage different sets of tables. But all users have to grant > insert, update, delete privileges to each other. There is an overhead > on DBA task obviously. > > Does the multiple users mechanism has any benefits on JDBC part of the > application? For example, a better connections? (There is a connection > pool) Without more information about what you are doing, the number of users, the needs for security, load balancing, app server, etc. all one can do is toss out random thoughts and hope they stick. Can you provide a bit of background? -- Daniel A. Morgan University of Washington damorgan@x.washington.edu (replace 'x' with 'u' to respond) Thank you for replying. I apol...

oracle high cpu
hi *, oracle 11, solaris 10. 4cpu machine 8GB mem/CPU. Oracle comes up to ~30% cpu from time to time and it stays there with that CPU usage untill i do init 6. Nothing I can thing of accounts to that usage (no sql queries at time it goes to 30% cpu) How can I check what accounts to this high cpu usage? On Thu, 07 Aug 2008 22:34:59 +0200, FrenKy <ime@prezime.com> wrote: >hi *, >oracle 11, solaris 10. 4cpu machine 8GB mem/CPU. > >Oracle comes up to ~30% cpu from time to time and it stays there with >that CPU usage untill i do init 6. >Nothing I can thing of accounts to that usage (no sql queries at time it >goes to 30% cpu) > >How can I check what accounts to this high cpu usage? 30 percent is way too low to pay attention to it. Oracle should run on a dedicated server. Apart from that the usual Solaris utilities apply, as you can query v$sysstat (Statistic descriptions are in v$statname) -- Sybrand Bakker Senior Oracle DBA sybrandb@hccnet.nl wrote: > On Thu, 07 Aug 2008 22:34:59 +0200, FrenKy <ime@prezime.com> wrote: > >> hi *, >> oracle 11, solaris 10. 4cpu machine 8GB mem/CPU. >> >> Oracle comes up to ~30% cpu from time to time and it stays there with >> that CPU usage untill i do init 6. >> Nothing I can thing of accounts to that usage (no sql queries at time it >> goes to 30% cpu) >> >> How can I check what accounts to this high cpu usage? > > 30 percent ...

high sys cpu on cpu# based on mpstat, but if using prstat -Lm -P 7, little cpu is used
The following prstat show little cpu is used on CPU 7# test$> prstat -Lm -P 7 PID USERNAME USR SYS TRP TFL DFL LCK SLP LAT VCX ICX SCL SIG PROCESS/LWPID 4020 oracle 6.0 1.3 0.0 0.0 0.0 2.4 90 0.2 2K 33 18K 0 mysqld/ 9 417 root 0.0 0.0 0.0 0.0 0.0 0.0 100 0.0 12 0 60 6 in.mpathd/1 4020 oracle 0.0 0.0 0.0 0.0 0.0 0.0 100 0.0 6 0 55 0 mysqld/ 1 1022 root 0.0 0.0 0.0 0.0 0.0 0.0 100 0.0 0 0 0 0 lcfd/1 843 root 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0 0 0 0 fmd/18 879 root 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0 0 0 0 syslogd/52...

testing individual cpu in multi-cpu system
anyone have any advice on tests of individual cpus in multi-cpu systems? what we'd like to do is be able to excercise one or the other cpu to see if we could force a failure. tks, jah -----= Posted via Newsfeeds.Com, Uncensored Usenet News =----- http://www.newsfeeds.com - The #1 Newsgroup Service in the World! -----== Over 100,000 Newsgroups - 19 Different Servers! =----- J. Holly wrote: > anyone have any advice on tests of individual cpus in multi-cpu systems? what > we'd like to do is be able to excercise one or the other cpu to see if we > could force a failure...

Diff of Sun Java System AppServer vs. SunONE Server vs. Sun Java System Web Server ?
What are the differences of these three Java Servers from Sun ? Ulf Sun Java System AppServer is the new name for SunONE Server. SunONE is no longer used by Sun it seems. The Sun Java System Web Server is just that a Web Server (also does JSP/Servlets) but no EJB. Ulf Meinhardt wrote: > What are the differences of these three Java Servers from Sun ? > > Ulf > ...

allocation cpu resources on multi cpu sun system
I am a student/EE attending a major university in southern california. My question is reguarding dynamic cpu resource allocation. Currently, we are limited to 12.5% of the total processing power at any given time. The system in question is a 13+ sparcs cpu sun system. When I use some of the chip design software (mentor graphics leonardo), it can take up to 4 hours to finish one task because I am limited to 12.5% cpu power at any given time while idle processes are upwards of 75%. I am up until 4am due to this poor cpu power allocation scheme. I would like to bring this issue up with the ...

high system CPU usage
This is on an idle system, a W2100z running an OpenSolaris build-24 kernel. (DEBUG). I think I had the same problem with the Nevada build that was on the box before. One of the CPUs is all bound up with system usage. I *know* this is a job for DTrace, but I'm not yet savvy enough to work out exactly what to do with it. (I've only previously studied the behaviour of applications.) Here's what the standard tools tell me. First vmstat: 0 0 0 2220160 1387472 0 73 0 0 0 0 0 0 0 0 0 26000 1436 25997 2 39 60 0 0 0 2220160 1387476 0 0 0 0 0 0 0 0 0 0 0 25945 1589 259...

oracle
Hi, Few days ago my database "eat" a lots of CPU and has high events waits and I/O waits. Now problem is gone, but I want to know what to do in future if that situation come back. How to find which session/sql cause problem? What to do with that session? Thanks in advance, kawu. On Jul 30, 10:16=A0am, kawu <k...@spam.pl> wrote: > Hi, > > Few days ago my database "eat" a lots of CPU and has high events waits > and I/O waits. Now problem is gone, but I want to know what to do in > future if that situation come back. How to find which session/sql cause > problem? What to do with that session? > > Thanks in advance, > kawu. What version and what platform are you running oracle on? Enterprise edition, standard, or what? What monitoring tools does your organization have setup if any? Do you have any kind of direct access into the database and what kind of role do you have? If your database is relatively recent OEM provides a fairly nice way to see what's impacting your database. Using it well depends on your experience level with oracle as a database and your experience as a developer/tuner. Well, you need a baseline to compare against when the peak occurs. Check out using STATSPACK, 15 min intervals so when the next peak happens you'll be able to see what's different. hpuxrac pisze: > On Jul 30, 10:16 am, kawu <k...@spam.pl> wrote: >> Hi, >> >> Few days ago my database "...

[9fans] High interruptload on cpu server
Moin, I have a cpu server running in a vmware machine. I run stats on my terminal (also vmware) and on the cpu server and since some days there is a relativ high interrupt/sec load on the cpu server. The stats line remains on 50% of the display. All other stats are "normal". The interrupt load on the terminal is much smaller. Is there any way to detect, how is eating the interrupts? There isn't much load on the cpu server. Matthias you can use kprof(3) to profile the kernel, usually the interrupt routine of the device driver for the device generating the interrupts will be on...

Responsiveness of Server at high CPU load
(Excerpt from a TAR - still open) From time to time, our Oracle test server (9.2.0.4 on Intel/Linux, 2 CPUs) got unusuable at CPU load of 99% as shown by top; in this state, nothing else could be done with Oracle, even trying to connect via sqlplus took about 1 hour (assuming one would wait that long). Processes running were Oracle processes and kswap (meaning that swapping was heavily taking place). Users complain in such a situation and my only remedy has been to reboot the server. pstack and oradebug could not be used. After analyzing lots of things we found out that nothing seems to be wrong with the database - it is just that a very inefficient query is running which blocks the Oracle server and avoids any other activity. Well, one message was found in the alert log, saying ksbsrv: No startup acknowledgement from forked process after 30 seconds but no ORA- error appears. Statspack Reports revealed a unusuable high "process startup" wait time. According to my experience under the Sun/Solaris platform, even if the 4 CPUs of our E3500 are at maximum load (showing an average idle of 0%), the Oracle (8.1.7) server is still available for new sessions (which run of course slower than usual). This happens quite often by the way, so it is a reliable experience. Assuming that the situation is caused by a bad query, I am concerned about the limited responsiveness of the server, since most of our queries are of batch type and run hours in the production platform, which...

Re: [9fans] High interruptload on cpu server
If you're on a PC, you can look to see how much time is spent in interrupts by 'acid -k $pid' and dumping the intrtimes array. Take a look at trap.c and you'll understand. It's just an array of 20 buckets for each possible interrupt level. Each bucket represents a 100 µ second interval. You'll be able to tell if any one level is hogging a lot of time. I've debated adding a user readable interface but then I'ld have to write a man page. kprof is moderately crappy for timing interrupts since it samples in the clock routine which is splhi'd as are th...

a database has very high accumulated CPU time for CKPT
oracle 4497646 1.3 0.0 60816 32964 - A Nov 18 911:51 ora_ckpt_prod oracle 4595964 0.0 0.0 53760 26020 - A Nov 18 20:33 ora_smon_prod oracle 4116580 0.0 0.0 57300 29496 - A Nov 18 14:37 ora_dbw0_prod oracle 4702444 0.0 0.0 55468 27624 - A Nov 18 10:49 ora_lgwr_prod oracle 4673758 0.0 0.0 56616 28848 - A Nov 18 5:06 ora_arc1_prod oracle 4509946 0.0 0.0 56616 28836 - A Nov 18 4:59 ora_arc0_prod what does this mean ? n...@hotmail.com wrote: > oracle 4497646 1.3 0.0 60816 32964 - A Nov 18 911:51 > ora_ckpt_prod > oracle 4595964 0.0 0.0 53760 26020 - A Nov 18 20:33 > ora_smon_prod > oracle 4116580 0.0 0.0 57300 29496 - A Nov 18 14:37 > ora_dbw0_prod > oracle 4702444 0.0 0.0 55468 27624 - A Nov 18 10:49 > ora_lgwr_prod > oracle 4673758 0.0 0.0 56616 28848 - A Nov 18 5:06 > ora_arc1_prod > oracle 4509946 0.0 0.0 56616 28836 - A Nov 18 4:59 > ora_arc0_prod > > what does this mean ? This forum is called comp.databases.oracle.server, not comp.databases.oracle.riddles. No version info, no init.ora info about incremental checkpointing, basically you are just *dumping* a ps display without even disclosing how long the database has been running? What do you expect? Do you need a clairvoyant or someone who does your work (for free)? -- Sybrand Bakker Senior Oracle DBA niy...

system idle process high cpu usage
hello I was having problems with 100% cpu usage & at the Microsoft site they specified to install the patch to overcome the problem & i did. Now System idle Process shows 100% cpu usage. Please give the solution at the earliest. Thanking you. -- matrex722 ...

'CPU used when call started' vs 'CPU used by this session'
Hello I have recently read somewhere that the statistics (from v$sesstat/mystat/sysstat) 'CPU used when call started' and 'CPU used by this session' are (nearly) equivalent but that 'CPU used when call started' should be used instead of 'CPU used by this session' because the latter is buggy. Now, is that true? And if so, why does Oracle have two statistics for the same thing. If not so, what is the difference. I have played a bit with it and indeed, the values seem to be identical for both statistics. Thanks, Rene -- Rene Nyffenegger http://www.adp-gmbh.ch/ ...

High system CPU on HP-UX with async io
We run Sybase 12.5.2 on HP-UX 11i. async io is on (mandatory in the installation guide.) We graph UNIX CPU detail and summary every day - system CPU is well above 10% whenever Sybase is busy. Looking at "glance", all the time is being wasted in select() and kill(). We have an equivalent spec system where async io is off (probably wrongly, since it is mandatory,) and system CPU is always below 10%. This really is a problem, because the async-on server is running slowly. I think if this could be corrected the server would run faster. I find it hard to believe this is correct behaviour and attribute the Server's slowness to this system CPU, which I think is wasted time. Does anyone have any experience of this problem and/or insight? Thanks for any help. Regards, Abe Probably, the extra CPU is expent checking for I/O completion. To confirm this, you may run a sp_sysmon report; it's expected to show a somewhat high level of I/O Busy, as compared with CPU Busy. To reduce disk I/O checks (along with network checks) you may increase the parameter "i/o polling process count", but beware this parameter has many trade-offs. By increasing it you may reduce CPU expent in I/O checks, but may also increase response time for I/O-bound applications. Please don't modify it before reading about it in the "System Administration Guide" Regards, Mariano Corral abe wrote: > We run Sybase 12.5.2 ...

Web resources about - High system cpu (vs. user cpu) - comp.databases.oracle.server

Type system - Wikipedia, the free encyclopedia
A type system associates a type with each computed value. By examining the flow of these values, a type system attempts to ensure or prove that ...

Chevy Silverado tows the line with trailering camera system
Filed under: Chevrolet , Truck Chevy is offering a new trailering camera system on its Silverado pickup, as well as a fifth-wheel prep pack, ...

Off-the-shelf BYOD systems bring privacy and security risks
When companies allow staff to use their own systems to access corporate data, the devices used can often be outside of IT department control. ...

Nissan & The World’s Largest Grid-Integrated Electric Vehicle System
... electric car technology to power the offices — in fact, it will be implementing the world’s largest grid-integrated electric vehicle (EV) system. ...

Square retailers rave about system
Jack Dorsey dishes on the ins-and-outs of financial service company Square.

Head of DC metro system announces 29-hour shutdown
WASHINGTON (AP) — The Latest on decision to shut down entire DC subway for inspections (all times local):

Are you SURE your DR systems are all working?
Flashback about a decade, to when this retail chain's data center needs a sprinkler head changed for its fire-suppression system which turns ...

Confusing multipronged Hyrule Warriors Season Pass system explained
... read them all if you're planning on pulling the trigger. Nintendo even has an FAQ at the end! In short, the Adventure maps and My Fairy system ...

Skyport Systems Lands $30M in VC to Secure Infrastructure
SkySecure is a remotely managed platform that establishes a secure enclave for an enterprise's most important business applications.

Cisco Systems, Inc. (CSCO) CMO Karen Walker Sells 6234 Shares of Stock
Cisco Systems, Inc. (CSCO) CMO Karen Walker Sells 6234 Shares of Stock Financial Market News Cisco Systems logo Cisco Systems, Inc. (NASDAQ:CSCO) ...

Resources last updated: 3/20/2016 2:28:43 AM