Solaris paging bringing system to halt


 
Thread Tools Search this Thread
Operating Systems Solaris Solaris paging bringing system to halt
# 1  
Old 01-30-2013
Java Solaris paging bringing system to halt

Hi experts,

I am running a solaris 10 with oracle 10 on a system with 96 GB physical memory. the system has a lot of swap space. the oracle SGA is 50 GB and PGA 40 GB. Here are the stats

Code:
 vmstat -i  

interrupt         total     rate
--------------------------------
clock         326481233      100
--------------------------------
Total         326481233      100

vmstat 5 10

 kthr      memory            page            disk          faults      cpu
 r b w   swap  free  re  mf pi po fr de sr m1 m1 m2 s0   in   sy   cs us sy id
 0 6 88 103634912 11344424 3220 2673 23858 1484 2478 1490560 7968 52 34 34 34 5285 56124 15227 21 7 72
 0 4 198 88671744 1665176 172 1878 4313 0 0 880176 0 143 71 69 69 5437 51868 16730 29 3 67
 0 1 198 88745880 1689648 265 2645 2167 2 2 755856 0 42 21 21 21 4277 52012 11342 31 3 66
 0 3 198 88650664 1590256 821 5144 8336 3933 8213 505112 95362 106 74 71 72 5540 53932 17424 29 5 66
 1 5 198 88551656 1590384 1257 7936 9253 12517 20882 1382704 32674 194 151 153 154 7152 65782 17224 32 7 61
 0 2 198 88648936 1744608 1212 7342 6150 6464 6464 920984 0 56 28 28 28 4886 61479 16384 28 5 67
 0 2 198 88646136 1722072 559 5884 6312 5 5 833776 0 90 45 45 45 4874 48578 13458 26 4 70

vmstat -s 5 5  
  
    30425 swap ins
      859 swap outs
    91275 pages swapped in
   434289 pages swapped out
8727192512 total address trans. faults taken
2183732165 page ins
187521189 page outs
9736382137 pages paged in
605506879 pages paged out
10512847480 total reclaims
10510567781 reclaims from free list
        0 micro (hat) faults
8727192512 minor (as) faults
879238109 major faults
2393782457 copy-on-write faults
3710680229 zero fill page faults
26013460920 pages examined by the clock daemon
     2104 revolutions of the clock hand
1011418148 pages freed by the clock daemon
 30859446 forks
   399501 vforks
 29685908 execs
49713902404 cpu context switches
17255648790 device interrupts
11140685959 traps
183233384578 system calls
32054311334 total name lookups (cache hits 100%)
3340616736 user   cpu
1062363299 system cpu
11268091303 idle   cpu

 vmstat -S 5 5
 kthr      memory            page            disk          faults      cpu
 r b w   swap  free  si  so pi po fr de sr m1 m1 m2 s0   in   sy   cs us sy id
 0 6 88 103634672 11344272 0 0 23858 1484 2478 115456 7968 52 34 34 34 5285 56124 15227 21 7 72
 0 3 198 88611128 1640584 0 0 10621 2 2 1201224 0 37 18 18 18 4716 44020 11982 23 5 73
 1 3 198 88696856 1717640 0 0 4550 0 0 962832 0 31 16 15 15 4347 33289 10654 19 3 78
 0 6 198 88646608 1650912 0 0 5869 0 0 612120 0 47 23 24 24 4993 51340 12713 28 3 69
 0 3 198 88739120 1736112 0 0 4602 2 2 495912 0 25 13 13 13 4175 69093 13916 20 4 76

 ipcs -m
IPC status from <running system> as of Wed Jan 30 17:52:26 EST 2013
T         ID      KEY        MODE        OWNER    GROUP
Shared Memory:
m 1761607788   0x60a9c490 --rw-r-----   odb10g      dba
m 1761607787   0          --rw-r-----   odb10g      dba
m 1761607786   0          --rw-r-----   odb10g      dba


sar -r  5 10

SunOS dc4dcmsdb02 5.10 Generic_147440-25 sun4u    01/30/2013

17:54:32 freemem freeswap
17:54:38  212340 177533677
17:54:43  202664 177614995
17:54:49  193226 177473616
17:54:54  207720 177643402
17:55:00  216316 177779968
17:55:06  214926 177775299
17:55:11  223668 177949459
17:55:17  215131 177828259
17:55:22  224671 178000899

echo ::memstat | mdb -k
Page Summary                Pages                MB  %Tot
------------     ----------------  ----------------  ----
Kernel                    3867118             30211   31%
Anon                      7513473             58699   61%
Exec and libs               43532               340    0%
Page cache                 626920              4897    5%
Free (cachelist)           199616              1559    2%
Free (freelist)            108101               844    1%

Total                    12358760             96552
Physical                 12349425             96479

can you help with what I need (besides adding more memory)? any paging parameters that I can tune?

thanks.

Last edited by jim mcnamara; 01-30-2013 at 11:51 PM..
# 2  
Old 01-31-2013
What's the output from:

Code:
echo ::memstat | mdb -k

That needs to be run by root.
# 3  
Old 01-31-2013
I posted it. the last command starting with "echo". And it is ran by root
# 4  
Old 04-15-2013
Did you resolve your issue yet?
# 5  
Old 04-16-2013
Hi Dustin,

the issue has been resolved by adding memory to the system.

Regards,
# 6  
Old 04-16-2013
Yes, that will usually do the trick.
Login or Register to Ask a Question

Previous Thread | Next Thread

10 More Discussions You Might Find Interesting

1. AIX

64K Paging taking place when available system ram exists

We are on AIX 6100-03-01 and our Admin said we are hitting "64K PAGING TAKING PLACE WHEN AVAILABLE SYSTEM RAM EXISTS " issue addressed by IZ71987 / IZ71850 / IZ72031 / IZ71191 / IZ71603. We are thinking to upgrade the OS to AIX 6.1 TL 6 SP3. Wondering if this issue "64K PAGING TAKING PLACE... (1 Reply)
Discussion started by: Sunusernewbie
1 Replies

2. Emergency UNIX and Linux Support

Shutdown to a Halt and restart

I've got critical patching this weekend on 6 HP UX machines, back to back and the customer has requested to shutdown, and halt the machines and to bring it up at a later period. The thing is the servers are located offshore and we are just the support team. I have a special login thru a console... (16 Replies)
Discussion started by: hedkandi
16 Replies

3. Homework & Coursework Questions

Four-Level multi-paging on x86 system with 64 bit addressing

1. The problem statement, all variables and given/known data: Hi all, I've got a huuuuuuge problem with understanding this new concept of multi-paging. I really tried to research but i could not find anything significant. I've been trying to understand this for 4 days and i cannot. The question... (0 Replies)
Discussion started by: snowboarder
0 Replies

4. SCO

SCO Open Server 6 - System Halt

Dear All, Newbie here, hope my post goes to the correct room. System: SCO Open Server 6.0 Everything running well, after this Sunday morning noone can login via telnet or via putty, the system suddenly reject without any message (and also ftp). The local IT Support there (without enough... (2 Replies)
Discussion started by: virgani
2 Replies

5. UNIX for Advanced & Expert Users

SCO Open Server 6 - System Halt

========================================================= Errrr previously post: https://www.unix.com/sco/140055-sco-open-server-6-system-halt.html Then I think it's better to post here. If I wrong, please ignore. ========================================================= Dear All, Newbie... (1 Reply)
Discussion started by: virgani
1 Replies

6. Solaris

unable to halt the zone

Hi Experts, I have to delete the non- global zone from one of the server. As per my understanding we need to halt the zone , then uninstall and later delete it with zoneadm however once I am trying to halt the zone the screen is getting hanged, after canceling the process and... (23 Replies)
Discussion started by: kumarmani
23 Replies

7. Solaris

difference b/w halt shutdown

Hi, I want to know the difference between halt and shutdown commands and which is better to use? Regards, visu (2 Replies)
Discussion started by: visu_ak
2 Replies

8. Solaris

rsh commands not getting executed from Solaris 10 System to AIX System

Hi Friends, I am trying to execute rsh commands from Solaris 10 system to AIX system. When I give; Solaris10# rsh <hostname> ls -l , it gives me an error rshd : 0826-826 The host name for your address is not known At the same time, Solaris10# rsh <hostname> ---- gives me remote shell of... (25 Replies)
Discussion started by: jumadhiya
25 Replies

9. Programming

How to halt Prog Execution for some time?

Hi, Perhaps I am asking a silly question, but I really don't know about it. Can anyone tell me the function for "sleep" kind of functionality in C language for Unix. I don't think any function with the name of sleep() exits in Unix's C language. or perhaps I am not known with the header file... (3 Replies)
Discussion started by: Ahsan
3 Replies

10. UNIX for Advanced & Expert Users

where's solaris halt log?

hi folks could tell my anyone where solaris 8 stores the logs about shutdown/halt command. I found no entries in syslog, but there more logs directories. probabaly a stupid question, but fast help would be great! thanks king regards, patrick (6 Replies)
Discussion started by: mrsaint
6 Replies
Login or Register to Ask a Question