Solved

when running mpstat or vmstat in solaris, sys is using the cpu alot more than usr.  How can I find out what the system is doing and why it is hogging up the resources, slowing the system down.

Posted on 2008-10-24
3
906 Views
Last Modified: 2013-12-27
I am running SQL Remote Sybase replication process on Solaris 10.  The hardware is a V880 with 8CPU and 32GB memory.  WHen the process gets to "writetext", mpstat shows that sys is using more of the cpu than usr, and the server is very slow, even to type simple commands.  How can I tell what the sys is doing?  Any ideas how I can track this?
0
Comment
Question by:Blademonkey
[X]
Welcome to Experts Exchange

Add your voice to the tech community where 5M+ people just like you are talking about what matters.

  • Help others & share knowledge
  • Earn cash & points
  • Learn & ask questions
3 Comments
 
LVL 11

Accepted Solution

by:
jgiordano earned 125 total points
ID: 22801034
Did you try using truss?

Or even dtrace since it is solaris 10?

here are some dtrace tools http://www.brendangregg.com/dtrace.html#Scripts

this dtrace script might be helpful - http://www.brendangregg.com/DTrace/lostcpu.html

0
 
LVL 1

Author Closing Comment

by:Blademonkey
ID: 31509866
Thank you very much!!
0
 
LVL 19

Expert Comment

by:grant300
ID: 22805030
Writetext usually means it is stuffing CLOBs or BLOBs into the database.  If you do not have any BLOBs or CLOBs in your schema, you can tune out now since this cannot be the problem.  The SYS time is all the driver level activity trying to get things to disk.  On Linux, you can see this as a bunch of I/O wait time.

Solaris is a very conservative operating system which is good for enterprise applications.  Unfortunately, it does some things in the name of stability that are not particularly helpful for database performance.  One of those is to, by default, turn off on-disk write caching.  This has an inordinately high impact on disk write performance, particularly for long sequential I/Os like file copies and continuous database writes to disk.

Fortunately, you can turn on-disk (the controller board bolted to the disk drive itself) caching at the Solaris driver level.  I don't recall the specifics but I have lived this problem before and the results are like night and day.  Definitely worth the time to track down the method for doing this.

BTW, ASE 12.5 was not all that stable.  v12.5.1 is the earliest release I would recommend.

Regards,
Bill
0

Featured Post

Independent Software Vendors: We Want Your Opinion

We value your feedback.

Take our survey and automatically be enter to win anyone of the following:
Yeti Cooler, Amazon eGift Card, and Movie eGift Card!

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

Suggested Solutions

I have been running these systems for a few years now and I am just very happy with them.   I just wanted to share the manual that I have created for upgrades and other things.  Oooh yes! FreeBSD makes me happy (as a server), no maintenance and I al…
Every server (virtual or physical) needs a console: and the console can be provided through hardware directly connected, software for remote connections, local connections, through a KVM, etc. This document explains the different types of consol…
This video shows how to set up a shell script to accept a positional parameter when called, pass that to a SQL script, accept the output from the statement back and then manipulate it in the Shell.
In a previous video, we went over how to export a DynamoDB table into Amazon S3.  In this video, we show how to load the export from S3 into a DynamoDB table.

696 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question