Login or Register to Ask a Question and Join Our Community


UNIX for Advanced & Expert Users

j2_maxPageReadAhead & j2_nBufferPerPagerDevice and CIO

Login to Discuss or Reply to this Discussion in Our Community

 
Thread Tools Search this Thread
Top Forums UNIX for Advanced & Expert Users j2_maxPageReadAhead & j2_nBufferPerPagerDevice and CIO
# 1  
Old 09-03-2008
j2_maxPageReadAhead & j2_nBufferPerPagerDevice and CIO
Hi Guys,

At the moment I'm experiencing a problem where one of the filesystems is being locked up because it looks like is running out of buffers and according to IBM a workaround should be decreasing the size of the
j2_maxPageReadAhead. My question is what if I turn on CIO at the filesystem level . I wonder whether enabling CIO then will bypass all these filesystems buffers such as j2_nBufferPerPagerDevice and j2_nBufferPerPagerDevice . Any thoughts?. Actually I forget to mention this is on AIX 5.3.

Harby.
# 2  
Old 09-03-2008
There is an AIX subforum here - post AIX problems there next time please^^

How did you notice it is locked up? Can you please post a vmstat -vs? An iostat of the disk(s) on which the problematic filesystems is placed on could help too.

Edit: Are you using AIO? If yes, "lsattr -El aio0" and "ps aux| grep aioserver" and iostat with -AQ to check aio queues maybe, please, ty Smilie What kind of application is causing that traffic?
Last edited by zaxxon; 09-03-2008 at 11:11 AM.. Reason: one more idea
# 3  
Old 09-03-2008
Hi there,

The below is the output of vmstat and iostat from the last ocurrence of this problem.

vmstat

r b avm fre re pi po fr sr cy in sy cs us sy id wa pc ec
1 5 1778357 24532 0 0 0 0 0 0 11 188 341 0 1 49 49 0.03 2.3
0 5 1778342 24547 0 0 0 0 0 0 29 407 378 0 3 50 47 0.05 4.4

iostat

tty: tin tout avg-cpu: % user % sys % idle % iowait physc % entc
0.0 323.2 0.6 1.8 49.3 48.3 0.0 3.1


thanks a lot for you reply.
# 4  
Old 09-04-2008
I asked for "vmstat -vs", not a simple vmstat. Also the iostat I asked for was for the disks which you think might make trouble.

The only thing I can see is that you have a high I/O wait.
# 5  
Old 09-04-2008
Actually this is all what I've got.
# 6  
Old 09-05-2008
When I type "vmstat -vs" on any AIX 5.3 machine, I get:

Code:
root@somebox:/> vmstat -vs
           1743435524 total address trans. faults
            400484880 page ins
             66634052 page outs
                    0 paging space page ins
                    0 paging space page outs
                    0 total reclaims
            896069655 zero filled pages faults
             79684162 executable filled pages faults
            969679308 pages examined by clock
                 1453 revolutions of the clock hand
            379300865 pages freed by the clock
             12301371 backtracks
                    0 free frame waits
                    0 extend XPT waits
              9308168 pending I/O waits
            459670082 start I/Os
             32358220 iodones
           1045168868 cpu context switches
            192183708 device interrupts
            106783424 software interrupts
            580781613 decrementer interrupts
               266715 mpc-sent interrupts
               215160 mpc-receive interrupts
            136135094 phantom interrupts
                    0 traps
           9578158818 syscalls
              4259840 memory pages
              4002637 lruable pages
               308553 free pages
                    6 memory pools
               569732 pinned pages
                 80.0 maxpin percentage
                  5.0 minperm percentage
                 80.0 maxperm percentage
                 64.4 numperm percentage
              2579734 file pages
                  0.0 compressed percentage
                    0 compressed pages
                 64.4 numclient percentage
                 80.0 maxclient percentage
              2579734 client pages
                    0 remote pageouts scheduled
                   28 pending disk I/Os blocked with no pbuf
                    0 paging space I/Os blocked with no psbuf
                 2484 filesystem I/Os blocked with no fsbuf
                  585 client filesystem I/Os blocked with no fsbuf
                 3554 external pager filesystem I/Os blocked with no fsbuf
                    0 Virtualized Partition Memory Page Faults
                 0.00 Time resolving virtualized partition memory page faults

So either you still did not enter that command or your vmstat is not recognizing parameters correct, which I doubt.
Login or Register to Ask a Question

Previous Thread | Next Thread

4 More Discussions You Might Find Interesting

1. UNIX for Advanced & Expert Users

CIO in AIX

Hi, I'm reading AIX documentation : Examples 1.To list the mounted file systems, enter: mountThis command produces output similar to the following: node mounted mounted over vfs date options ---- ------- ------------ --- ------------ ... (2 Replies)
Discussion started by: big123456
2 Replies

2. Linux

Is Concurrent I/O (CIO) available on Linux?

Hi Guys, I just wondering if any of the AIX gurus with some exposure to Linux Redhat can tell me whether in Linux ( Red Hat or OpenSuse ) we have a similar option/capability like the CIO (Concurrent I/O) which is currently available on AIX. The reason I'm asking is because in the past we... (2 Replies)
Discussion started by: arizah
2 Replies

3. AIX

turning CIO on and how to monitor

Hi Guys, I have a database server where we run AIX 5.3 on a power5 box and we just turned on CIO (concurrent I/O) for the database filesystems. Now my assumption is that enabling CIO the database basically will bypass the filesystem cache releasing some extra memory that can be allocated... (1 Reply)
Discussion started by: hariza
1 Replies

4. UNIX for Advanced & Expert Users

turning CIO on and how to monitor

Hi Guys, I have a database server where we run AIX 5.3 on a power5 box and we just turned on CIO (concurrent I/O) for the database filesystems. Now my assumption is that enabling CIO the database basically will bypass the filesystem cache releasing some extra memory that can be allocated... (1 Reply)
Discussion started by: hariza
1 Replies
Login or Register to Ask a Question