From: Jim Lane (jimlane@SYMPATICO.CA)
Date: Thu May 22 2003 - 20:20:54 EDT
I don't know about Domino but I've used minperm and maxperm with good effect
on Oracle workloads and I'd suggest that 2 & 4 might be a bit extreme. To
start with I'd try something like
/usr/samples/kernel/vmtune -p10 -P30
and go lower if necessary. In the case of Oracle the effect of this hack shows
up within minutes, not hours, so it's easy to play with until you get it
right.
Regards,
Jim Lane
On May 22, 2003 10:59 am, Holger.VanKoll@SWISSCOM.COM wrote:
> I would try
>
> /usr/samples/kernel/vmtune -p2 -P4
>
> I can take hours before this could improve performance.
> But you should see much less po.
>
>
>
> -----Original Message-----
> From: abdellah.djebli@SICKKIDS.CA [mailto:abdellah.djebli@SICKKIDS.CA]
> Sent: Thursday, May 22, 2003 4:35 PM
> To: aix-l@Princeton.EDU
> Subject: Re: performance question
>
>
> Hi,
>
> Bobby, Thanks for the reply.
>
>
> Abdu
>
>
> NotesMail02:~ # lsps -a
> Page Space Physical Volume Volume Group Size %Used Active Auto
> Type
> paging00 vpath0 mail02vg 2944MB 34 yes no
> lv
> hd6 hdisk0 rootvg 2944MB 37 yes yes
> lv
>
> NotesMail02:~ # vmstat 2 10
> kthr memory page faults cpu
> ----- ----------- ------------------------ ------------ -----------
> r b avm fre re pi po fr sr cy in sy cs us sy id wa
> 2 2 708830 327 0 20 21 388 1319 0 4009 22088 7639 7 4 69 20
> 0 6 708834 622 0 157 286 897 3139 0 5059 99424 8152 12 11 30 48
> 3 2 708835 585 0 107 327 642 3019 0 4980 13371 8811 10 5 26 59
> 0 3 708835 320 0 174 308 1031 3181 0 5272 38763 8723 27 8 12 53
> 3 2 708835 390 0 200 0 1285 3214 0 4974 50426 8755 22 8 32 38
> 1 1 708836 324 0 104 0 385 915 0 4587 42178 8747 12 4 55 28
> 4 2 708836 658 0 59 53 770 1967 0 4727 64247 8708 20 6 38 36
> 6 1 708836 256 0 120 41 773 1963 0 4884 122609 8943 20 10 15 55
> 6 1 708836 383 0 21 170 514 1369 0 4724 33722 8872 6 6 49 39
> 0 0 708836 391 0 58 100 385 942 0 4684 109538 8803 14 8 53 25
>
>
> NotesMail02:~ # /usr/samples/kernel/vmtune
> vmtune: current values:
> -p -P -r -R -f -F -N
> -W
> minperm maxperm minpgahead maxpgahead minfree maxfree pd_npages
> maxrandwrt
> 147089 490299 2 8 240 496 65536
> 0
>
> -M -w -k -c -b -B -u -l
> -d
> maxpin npswarn npskill numclust numfsbufs hd_pbuf_cnt lvm_bufcnt
> lrubucket defps
> 838852 47104 11776 1 196 320 9 131072
> 1
>
> -s -n -S -L -g
> -h
> sync_release_ilock nokilluid v_pinshm lgpg_regions lgpg_size
> strict_maxperm
> 0 0 0 0 0 0
>
> -t -j -J -z
> maxclient j2_nPagesPer j2_maxRandomWrite j2_nRandomCluster
> 490299 32 0 0
>
> -Z -q -Q -y
> j2_nBufferPer j2_minPageReadAhead j2_maxPageReadAhead
> memory_affinity
> 512 2 8 0
>
> -V -i
> num_spec_dataseg spec_dataseg_int
> 0 512
>
> PTA balance threshold percentage = 50.0%
>
> number of valid memory pages = 1048565 maxperm=50.0% of real memory
> maximum pinable=80.0% of real memory minperm=15.0% of real memory
> number of file memory pages = 489001 numperm=49.9% of real memory
> number of compressed memory pages = 0 compressed=0.0% of real memory
> number of client memory pages = 0 numclient=0.0% of real memory
> # of remote pgs sched-pageout = 0 maxclient=50.0% of real memory
>
>
>
>
>
> Bob.Kelley@BRINKS
> INC.COM To:
> aix-l@Princeton.EDU
> Sent by: IBM AIX cc:
> Discussion List Subject: Re: performance
> question
> <aix-l@Princeton.
> EDU>
>
>
> 05/22/2003 09:36
> AM
> Please respond to
> IBM AIX
> Discussion List
>
>
>
>
>
>
> Hi Adbu,
>
> Can you please send a cut/paste of vmstat 2 10, vmtune, lsps -a???
>
>
> Bobby Kelley
> Sr. AIX/UNIX Administrator
> BRINKS
> 972-877-5341
>
> |---------+---------------------------->
> |
> | | abdellah.djebli@S|
> | | ICKKIDS.CA |
> | | Sent by: IBM AIX |
> | | Discussion List |
> | | <aix-l@Princeton.|
> | | EDU> |
> | |
> | |
> | | 05/22/2003 08:07 |
> | | AM |
> | | Please respond to|
> | | IBM AIX |
> | | Discussion List |
> |
> |---------+---------------------------->
>
> ------------------------------------------------------------------------
> ------------------------------------------------------|
>
> | To: aix-l@Princeton.EDU
> |
> | cc:
> |
> | Subject: performance question
>
> ------------------------------------------------------------------------
> ------------------------------------------------------|
>
>
>
>
>
> Hi All,
>
> We have moved a partitioned Domino Mail server from an SP wide silver
> node with ( 4 x 332 MHz and 3 GB RAM, AIX 4.3.3-ML10) to a P630 with ( 2
> x 1.45 GHz and 4 GB RAM, AIX-5.1 ML3, 64 bit kernel). In both cases the
> server is attached to a SAN with two 2GB fiber adapters. Although the
> response time is better, we see the same amount of paging and the same
> amount of IO wait, which in peak times reaches 40% however, nmon and/or
> iostat does not show a very high IO activity.
>
> Beside adding more RAM, anyone has some suggestions.
> Anyone has a Domino server running on P630.
>
> Thanks in advance.
>
> Abdu
>
> NOTICE TO BRINK'S EMPLOYEES:
> All electronic mail communications composed, sent, received or stored on
> or through Brink's equipment are subject to the Brink's Electronic
> Communications Policy. Employees should have no expectation of privacy
> in their use of the Company's electronic mail system.
This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 22:16:51 EDT