SUMMARY: runaway swap

From: <Dave_Pigliavento_at_mapinfo.com>
Date: Fri Nov 17 2006 - 09:46:08 EST
Thanks to Darren Dunham and Scott Lawson for there quick response and 
assistance determining the root cause.

Reviewing ::memstat indicated that the kernel was using 60% of the 
available memory. zfs will use large amounts of RAM if available to 
increase filesystem performance. In addition there are several bugs posted 
on opensolaris surrounding zfs memory usage behavior.

http://bugs.opensolaris.org/bugdatabase/view_bug.do;jsessionid=b1e7e259c94e62c1892e7684891:oZFF?bug_id=6416757

zfs should release this memory once other processes require the additional 
resources. 

Thanks again,
Dave



----- Forwarded by Dave Pigliavento/MapInfo Corp on 11/17/2006 09:34 AM 
-----

Dave_Pigliavento@mapinfo.com 
Sent by: sunmanagers-bounces@sunmanagers.org
11/16/2006 05:47 PM

To
<sunmanagers@sunmanagers.org>
cc

Subject
runaway swap






I have a new install of Solaris 10 6/06 on a sunfire v440. This system has 

4 newly configured zones which are not currently running any applications. 

Oracle is being installed on several of them but no database has been 
brought online yet. The swap partition is 32Gig in size and the system has 

16Gig of RAM.

During the course of an 8 hour day memory is depleted on the system. It is 

not gradual, instead large amounts of memory disappear at once. Shutting 
down all zones does not free up any of the memory and I can't determine 
who is using it. The global zone is not running any applications aside 
from the base install and some emc components for the SAN disk. The only 
way I have been able to regain the memory is to reboot the system.

There are only a few very small files located within /tmp or /var/run. Our 

dba's have installed oracle on 2 of the zones which I have to assume is 
what has caused this to happen 2 days in a row but I can't for the life of 

me figure out why.

Any assistance would be greatly appreciated.
Thanks,
Dave


root@myriad> swap -s
total: 151120k bytes allocated + 27168k reserved = 178288k used, 35891512k 

available

root@myriad> swap -l
swapfile             dev  swaplo blocks   free
/dev/md/dsk/d40     85,40     16 65543600 65543600

root@myriad> df -k
swap                 35890096     416 35889680     1%    /tmp
swap                 35889736      56 35889680     1%    /var/run

root@myriad> vmstat 5
 kthr      memory            page            disk          faults      cpu
 r b w   swap  free  re  mf pi po fr de sr m0 m1 m2 m1   in   sy   cs us 
sy id
 0 0 0 45302368 14381664 34 118 118 0 0 0 1 0  0  0  0  446 1013  735  0 1 

99
 0 0 0 35891152 4267200 10 91 0 0  0  0  0  6  6  6  4  282  187  182  0 0 

99

prstat -s rss
   PID USERNAME  SIZE   RSS STATE  PRI NICE      TIME  CPU PROCESS/NLWP
   536 root       16M 7368K sleep    1    0   0:00:10 0.0% naviagent/2
 18000 root     7824K 5656K cpu0     1    0   0:00:00 0.0% prstat/1
 11819 root     9680K 4200K sleep    1    0   0:00:05 0.0% svc.configd/13
     9 root     9816K 4192K sleep   29    0   0:00:06 0.0% svc.configd/14

prstat -Z
ZONEID    NPROC  SIZE   RSS MEMORY      TIME  CPU ZONE
     0       55  255M   80M   0.4%   0:01:17 0.0% global
     3       44  191M   45M   0.2%   0:00:30 0.0% solo
     4       42  173M   44M   0.2%   0:01:28 0.0% boonville
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers
Received on Fri Nov 17 09:48:05 2006

This archive was generated by hypermail 2.1.8 : Thu Mar 03 2016 - 06:44:03 EST