Extra Shared Memory MESSAGE Segments
IDS 11.50.FC5 AIX 5.3
I am trying to figure out why the Message Segment decided to split itself into
19 smaller pieces ... and if there is anything wrong with it being in so many
pieces or if I should strive to put them, ala Humpty Dumpty, back together
again ... and how?
Thanks in advance.
Clifton
Online.log:
Requested shared memory segment size rounded from 548KB to 560KB
03:13:34 Dynamically allocated new message shared memory segment (size 560KB)
03:13:34 Memory sizes:resident:2761408 KB, virtual:3201120 KB, no SHMTOTAL
limit
$ONCONFIG Parameters:
VPCLASS cpu,num=19,aff=(1-9),noage
NETTYPE ipcshm,19,50,CPU # corresponding nettype for cpu vpclass
NETTYPE soctcp,10,250,NET # no need for corresponding vpclass
-> onstat -g seg
IBM Informix Dynamic Server Version 11.50.FC5 -- On-Line -- Up 11:00:10 --5972048 Kbytes
Segment Summary:
id key addr size ovhd class blkused blkfree
1048579 52564801 700000010000000 2827681792 33570328 R* 690349 3
1048578 52564802 7000000c0000000 3276800000 38401840 V 581789 218211
1048577 52564803 700000190000000 573440 7984 M 140 0
4 52564804 7000001a0000000 573440 7984 M 140 0
5 52564805 7000001b0000000 573440 7984 M 140 0
6 52564806 7000001c0000000 573440 7984 M 139 1
7 52564807 7000001d0000000 573440 7984 M 135 5
8 52564808 7000001e0000000 573440 7984 M 135 5
9 52564809 7000001f0000000 573440 7984 M 135 5
10 5256480a 700000200000000 573440 7984 M 135 5
11 5256480b 700000210000000 573440 7984 M 135 5
12 5256480c 700000220000000 573440 7984 M 135 5
13 5256480d 700000230000000 573440 7984 M 135 5
14 5256480e 700000240000000 573440 7984 M 135 5
15 5256480f 700000250000000 573440 7984 M 135 5
16 52564810 700000260000000 573440 7984 M 135 5
17 52564811 700000270000000 573440 7984 M 135 5
18 52564812 700000280000000 573440 7984 M 135 5
19 52564813 700000290000000 573440 7984 M 135 5
20 52564814 7000002a0000000 573440 7984 M 135 5
21 52564815 7000002b0000000 573440 7984 M 135 5
Total: - - 6115377152 - - 1274722 218290
(* segment locked in memory)
-> onstat -g glo
IBM Informix Dynamic Server Version 11.50.FC5 -- On-Line -- Up 10:58:52 --5972048 Kbytes
MT global info:
sessions threads vps lngspins
2017 2152 38 0
sched calls thread switches yield 0 yield n yield forever
total: 146123725 64660569 87511186 1953442 8765693
per sec: 7534 7279 252 102 149
Virtual processor summary:
class vps usercpu syscpu total
cpu 19 5979.75 947.66 6927.41
aio 5 1.08 5.73 6.81
lio 1 0.08 0.20 0.28
pio 1 0.08 0.21 0.29
adm 1 3.71 4.77 8.48
soc 10 113.13 320.73 433.86
msc 1 12.84 4.22 17.06
total 38 6110.67 1283.52 7394.19
Individual virtual processors:
vp pid class usercpu syscpu total Thread Eff
1 78810 cpu 2219.96 381.83 2601.79 4272.51 60%
2 95468 adm 3.71 4.77 8.48 0.00 0%
3 86220 cpu 2226.48 333.36 2559.84 4282.16 59%
4 53908 cpu 880.21 151.26 1031.47 1744.49 59%
5 90588 cpu 304.57 52.11 356.68 559.29 63%
6 82910 cpu 141.73 12.81 154.54 209.78 73%
7 50262 cpu 64.07 6.04 70.11 97.68 71%
8 102636 cpu 45.59 3.34 48.93 71.90 68%
9 94688 cpu 55.71 2.66 58.37 73.51 79%
10 87010 cpu 31.81 1.69 33.50 45.98 72%
11 91360 cpu 7.39 0.60 7.99 11.95 66%
12 49834 cpu 0.64 0.30 0.94 3.37 27%
13 98788 cpu 0.62 0.27 0.89 2.26 39%
14 91110 cpu 0.18 0.21 0.39 1.98 19%
15 99562 cpu 0.11 0.18 0.29 0.35 83%
16 106714 cpu 0.12 0.20 0.32 0.58 55%
17 102888 cpu 0.10 0.18 0.28 0.28 100%
18 95210 cpu 0.11 0.18 0.29 0.29 100%
19 103660 cpu 0.11 0.18 0.29 0.29 100%
20 57862 cpu 0.24 0.26 0.50 2.25 22%
21 99310 lio 0.08 0.20 0.28 0.28 100%
22 107760 pio 0.08 0.21 0.29 0.29 100%
23 111858 aio 0.56 2.78 3.34 12.72 26%
24 115956 msc 12.84 4.22 17.06 17.31 98%
25 120054 aio 0.25 2.32 2.57 2.63 97%
26 124152 soc 11.79 32.80 44.59 NA NA
27 107002 soc 8.14 22.95 31.09 NA NA
28 103420 soc 28.65 90.19 118.84 NA NA
29 110846 soc 4.96 12.82 17.78 NA NA
30 128000 soc 9.25 25.22 34.47 NA NA
31 110850 soc 7.07 19.29 26.36 NA NA
32 107268 soc 6.57 16.66 23.23 NA NA
33 132102 soc 6.85 19.72 26.57 NA NA
34 114696 soc 7.56 20.80 28.36 NA NA
35 114954 soc 22.29 60.28 82.57 NA NA
36 136206 aio 0.10 0.23 0.33 0.33 100%
37 165034 aio 0.10 0.23 0.33 1.10 30%
38 177362 aio 0.07 0.17 0.24 1.06 22%
tot 6110.67 1283.52 7394.19
_________________________________________________________________
New Windows 7: Find the right PC for you.
http://www.microsoft.com/windows/pc-scout/default.aspx?CBID=wl&ocid=PID24727::T:
WLMTAGL:ON:WL:en-US:WWL_WIN_pcscout:112009
Are you seeing a performance issue related to shm connections? You
specified 19 segments in the onconfig by using NETTYPE ipcshm,19,50,CPU=
.
This will create 19 message segments with space for 50 connections each=
.
You can combine them a bit if you want, and that might be an advantage =
so
that you reduce the total number of memory segments. However, in order=
to
do that and have the same capacity, you would need to increase the size=
of
each segment.
=
From: "Clifton Bean" <clifton_bean@hotmail.com> =
=
To: ids@iiug.org =
=
Date: 11/02/2009 01:31 PM =
=
Subject: Extra Shared Memory MESSAGE Segments [17870] =
=
Sent by: ids-bounces@iiug.org =
=
IDS 11.50.FC5 AIX 5.3
I am trying to figure out why the Message Segment decided to split itse=
lf
into
19 smaller pieces ... and if there is anything wrong with it being in s=
o
many
pieces or if I should strive to put them, ala Humpty Dumpty, back toget=
her
again ... and how?
Thanks in advance.
Clifton
Online.log:
Requested shared memory segment size rounded from 548KB to 560KB
03:13:34 Dynamically allocated new message shared memory segment (size
560KB)
03:13:34 Memory sizes:resident:2761408 KB, virtual:3201120 KB, no SHMTO=
TAL
limit
$ONCONFIG Parameters:
VPCLASS cpu,num=3D19,aff=3D(1-9),noage
NETTYPE ipcshm,19,50,CPU # corresponding nettype for cpu vpclass
NETTYPE soctcp,10,250,NET # no need for corresponding vpclass
-> onstat -g seg
IBM Informix Dynamic Server Version 11.50.FC5 -- On-Line -- Up 11:00:10=
--5972048 Kbytes
Segment Summary:
id key addr size ovhd class blkused blkfree
1048579 52564801 700000010000000 2827681792 33570328 R* 690349 3
1048578 52564802 7000000c0000000 3276800000 38401840 V 581789 218211
1048577 52564803 700000190000000 573440 7984 M 140 0
4 52564804 7000001a0000000 573440 7984 M 140 0
5 52564805 7000001b0000000 573440 7984 M 140 0
6 52564806 7000001c0000000 573440 7984 M 139 1
7 52564807 7000001d0000000 573440 7984 M 135 5
8 52564808 7000001e0000000 573440 7984 M 135 5
9 52564809 7000001f0000000 573440 7984 M 135 5
10 5256480a 700000200000000 573440 7984 M 135 5
11 5256480b 700000210000000 573440 7984 M 135 5
12 5256480c 700000220000000 573440 7984 M 135 5
13 5256480d 700000230000000 573440 7984 M 135 5
14 5256480e 700000240000000 573440 7984 M 135 5
15 5256480f 700000250000000 573440 7984 M 135 5
16 52564810 700000260000000 573440 7984 M 135 5
17 52564811 700000270000000 573440 7984 M 135 5
18 52564812 700000280000000 573440 7984 M 135 5
19 52564813 700000290000000 573440 7984 M 135 5
20 52564814 7000002a0000000 573440 7984 M 135 5
21 52564815 7000002b0000000 573440 7984 M 135 5
Total: - - 6115377152 - - 1274722 218290
(* segment locked in memory)
-> onstat -g glo
IBM Informix Dynamic Server Version 11.50.FC5 -- On-Line -- Up 10:58:52=
--5972048 Kbytes
MT global info:
sessions threads vps lngspins
2017 2152 38 0
sched calls thread switches yield 0 yield n yield forever
total: 146123725 64660569 87511186 1953442 8765693
per sec: 7534 7279 252 102 149
Virtual processor summary:
class vps usercpu syscpu total
cpu 19 5979.75 947.66 6927.41
aio 5 1.08 5.73 6.81
lio 1 0.08 0.20 0.28
pio 1 0.08 0.21 0.29
adm 1 3.71 4.77 8.48
soc 10 113.13 320.73 433.86
msc 1 12.84 4.22 17.06
total 38 6110.67 1283.52 7394.19
Individual virtual processors:
vp pid class usercpu syscpu total Thread Eff
1 78810 cpu 2219.96 381.83 2601.79 4272.51 60%
2 95468 adm 3.71 4.77 8.48 0.00 0%
3 86220 cpu 2226.48 333.36 2559.84 4282.16 59%
4 53908 cpu 880.21 151.26 1031.47 1744.49 59%
5 90588 cpu 304.57 52.11 356.68 559.29 63%
6 82910 cpu 141.73 12.81 154.54 209.78 73%
7 50262 cpu 64.07 6.04 70.11 97.68 71%
8 102636 cpu 45.59 3.34 48.93 71.90 68%
9 94688 cpu 55.71 2.66 58.37 73.51 79%
10 87010 cpu 31.81 1.69 33.50 45.98 72%
11 91360 cpu 7.39 0.60 7.99 11.95 66%
12 49834 cpu 0.64 0.30 0.94 3.37 27%
13 98788 cpu 0.62 0.27 0.89 2.26 39%
14 91110 cpu 0.18 0.21 0.39 1.98 19%
15 99562 cpu 0.11 0.18 0.29 0.35 83%
16 106714 cpu 0.12 0.20 0.32 0.58 55%
17 102888 cpu 0.10 0.18 0.28 0.28 100%
18 95210 cpu 0.11 0.18 0.29 0.29 100%
19 103660 cpu 0.11 0.18 0.29 0.29 100%
20 57862 cpu 0.24 0.26 0.50 2.25 22%
21 99310 lio 0.08 0.20 0.28 0.28 100%
22 107760 pio 0.08 0.21 0.29 0.29 100%
23 111858 aio 0.56 2.78 3.34 12.72 26%
24 115956 msc 12.84 4.22 17.06 17.31 98%
25 120054 aio 0.25 2.32 2.57 2.63 97%
26 124152 soc 11.79 32.80 44.59 NA NA
27 107002 soc 8.14 22.95 31.09 NA NA
28 103420 soc 28.65 90.19 118.84 NA NA
29 110846 soc 4.96 12.82 17.78 NA NA
30 128000 soc 9.25 25.22 34.47 NA NA
31 110850 soc 7.07 19.29 26.36 NA NA
32 107268 soc 6.57 16.66 23.23 NA NA
33 132102 soc 6.85 19.72 26.57 NA NA
34 114696 soc 7.56 20.80 28.36 NA NA
35 114954 soc 22.29 60.28 82.57 NA NA
36 136206 aio 0.10 0.23 0.33 0.33 100%
37 165034 aio 0.10 0.23 0.33 1.10 30%
38 177362 aio 0.07 0.17 0.24 1.06 22%
tot 6110.67 1283.52 7394.19
_________________________________________________________________
New Windows 7: Find the right PC for you.
http://www.microsoft.com/windows/pc-scout/default.aspx?CBID=3Dwl&ocid=3D=
PID24727::T:WLMTAGL:ON:WL:en-US:WWL_WIN_pcscout:112009...
***********************************************************************=
********
Forum Note: Use "Reply" to post a response in the discussion forum.
=
Cliff,
Basically the engine creates a message segment for each poll thread and you
have configured 19 threads (one in each CPU VP), as Madison pointed out, so
19 shared memory segments. No real problem, even on HPUX/PA-RISC, as each
CPU VP only has to connect to the virtual segment, the resident segment
(unless these two have been combined by setting RESIDENT -1), and one
message segment. That still leaves at least one shared memory base address
register free for a second virtual segment if the engine needs to allocate
one without thrashing those registers (only a problem on PA-RISC anyway).
Art
Art S. Kagel
Oninit (www.oninit.com)
IIUG Board of Directors (art@iiug.org)
Disclaimer: Please keep in mind that my own opinions are my own opinions and
do not reflect on my employer, Oninit, the IIUG, nor any other organization
with which I am associated either explicitly or implicitly. Neither do
those opinions reflect those of other individuals affiliated with any entity
with which I am affiliated nor those of the entities themselves.
On Mon, Nov 2, 2009 at 3:17 PM, Madison Pruet <mpruet@us.ibm.com> wrote:
> Are you seeing a performance issue related to shm connections? You
> specified 19 segments in the onconfig by using NETTYPE ipcshm,19,50,CPU=
> ..
> This will create 19 message segments with space for 50 connections each=
> ..
>
> You can combine them a bit if you want, and that might be an advantage =
> so
> that you reduce the total number of memory segments. However, in order=
> to
> do that and have the same capacity, you would need to increase the size=
> of
> each segment.
>
> =
>
> From: "Clifton Bean" <clifton_bean@hotmail.com> =
>
> =
>
> To: ids@iiug.org =
>
> =
>
> Date: 11/02/2009 01:31 PM =
>
> =
>
> Subject: Extra Shared Memory MESSAGE Segments [17870] =
>
> =
>
> Sent by: ids-bounces@iiug.org =
>
> =
>
> IDS 11.50.FC5 AIX 5.3
>
> I am trying to figure out why the Message Segment decided to split itse=
> lf
> into
> 19 smaller pieces ... and if there is anything wrong with it being in s=
> o
> many
> pieces or if I should strive to put them, ala Humpty Dumpty, back toget=
> her
> again ... and how?
>
> Thanks in advance.
>
> Clifton
>
> Online.log:
>
> Requested shared memory segment size rounded from 548KB to 560KB
> 03:13:34 Dynamically allocated new message shared memory segment (size
> 560KB)
> 03:13:34 Memory sizes:resident:2761408 KB, virtual:3201120 KB, no SHMTO=
> TAL
> limit
>
> $ONCONFIG Parameters:
>
> VPCLASS cpu,num=3D19,aff=3D(1-9),noage
>
> NETTYPE ipcshm,19,50,CPU # corresponding nettype for cpu vpclass
> NETTYPE soctcp,10,250,NET # no need for corresponding vpclass>
> -> onstat -g seg
>
> IBM Informix Dynamic Server Version 11.50.FC5 -- On-Line -- Up 11:00:10=
> --> 5972048 Kbytes
>
> Segment Summary:
> id key addr size ovhd class blkused blkfree
> 1048579 52564801 700000010000000 2827681792 33570328 R* 690349 3
> 1048578 52564802 7000000c0000000 3276800000 38401840 V 581789 218211
> 1048577 52564803 700000190000000 573440 7984 M 140 0
> 4 52564804 7000001a0000000 573440 7984 M 140 0
> 5 52564805 7000001b0000000 573440 7984 M 140 0
> 6 52564806 7000001c0000000 573440 7984 M 139 1
> 7 52564807 7000001d0000000 573440 7984 M 135 5
> 8 52564808 7000001e0000000 573440 7984 M 135 5
> 9 52564809 7000001f0000000 573440 7984 M 135 5
> 10 5256480a 700000200000000 573440 7984 M 135 5
> 11 5256480b 700000210000000 573440 7984 M 135 5
> 12 5256480c 700000220000000 573440 7984 M 135 5
> 13 5256480d 700000230000000 573440 7984 M 135 5
> 14 5256480e 700000240000000 573440 7984 M 135 5
> 15 5256480f 700000250000000 573440 7984 M 135 5
> 16 52564810 700000260000000 573440 7984 M 135 5
> 17 52564811 700000270000000 573440 7984 M 135 5
> 18 52564812 700000280000000 573440 7984 M 135 5
> 19 52564813 700000290000000 573440 7984 M 135 5
> 20 52564814 7000002a0000000 573440 7984 M 135 5
> 21 52564815 7000002b0000000 573440 7984 M 135 5
> Total: - - 6115377152 - - 1274722 218290
>
> (* segment locked in memory)
>
> -> onstat -g glo
>
> IBM Informix Dynamic Server Version 11.50.FC5 -- On-Line -- Up 10:58:52=
> --> 5972048 Kbytes
>
> MT global info:
> sessions threads vps lngspins
> 2017 2152 38 0
>
> sched calls thread switches yield 0 yield n yield forever
> total: 146123725 64660569 87511186 1953442 8765693
> per sec: 7534 7279 252 102 149
>
> Virtual processor summary:
> class vps usercpu syscpu total
> cpu 19 5979.75 947.66 6927.41
> aio 5 1.08 5.73 6.81
> lio 1 0.08 0.20 0.28
> pio 1 0.08 0.21 0.29
> adm 1 3.71 4.77 8.48
> soc 10 113.13 320.73 433.86
> msc 1 12.84 4.22 17.06
> total 38 6110.67 1283.52 7394.19
>
> Individual virtual processors:
> vp pid class usercpu syscpu total Thread Eff
> 1 78810 cpu 2219.96 381.83 2601.79 4272.51 60%
> 2 95468 adm 3.71 4.77 8.48 0.00 0%
> 3 86220 cpu 2226.48 333.36 2559.84 4282.16 59%
> 4 53908 cpu 880.21 151.26 1031.47 1744.49 59%
> 5 90588 cpu 304.57 52.11 356.68 559.29 63%
> 6 82910 cpu 141.73 12.81 154.54 209.78 73%
> 7 50262 cpu 64.07 6.04 70.11 97.68 71%
> 8 102636 cpu 45.59 3.34 48.93 71.90 68%
> 9 94688 cpu 55.71 2.66 58.37 73.51 79%
> 10 87010 cpu 31.81 1.69 33.50 45.98 72%
> 11 91360 cpu 7.39 0.60 7.99 11.95 66%
> 12 49834 cpu 0.64 0.30 0.94 3.37 27%
> 13 98788 cpu 0.62 0.27 0.89 2.26 39%
> 14 91110 cpu 0.18 0.21 0.39 1.98 19%
> 15 99562 cpu 0.11 0.18 0.29 0.35 83%
> 16 106714 cpu 0.12 0.20 0.32 0.58 55%
> 17 102888 cpu 0.10 0.18 0.28 0.28 100%
> 18 95210 cpu 0.11 0.18 0.29 0.29 100%
> 19 103660 cpu 0.11 0.18 0.29 0.29 100%
> 20 57862 cpu 0.24 0.26 0.50 2.25 22%
> 21 99310 lio 0.08 0.20 0.28 0.28 100%
> 22 107760 pio 0.08 0.21 0.29 0.29 100%
> 23 111858 aio 0.56 2.78 3.34 12.72 26%
> 24 115956 msc 12.84 4.22 17.06 17.31 98%
> 25 120054 aio 0.25 2.32 2.57 2.63 97%
> 26 124152 soc 11.79 32.80 44.59 NA NA
> 27 107002 soc 8.14 22.95 31.09 NA NA
> 28 103420 soc 28.65 90.19 118.84 NA NA
> 29 110846 soc 4.96 12.82 17.78 NA NA
> 30 128000 soc 9.25 25.22 34.47 NA NA
> 31 110850 soc 7.07 19.29 26.36 NA NA
> 32 107268 soc 6.57 16.66 23.23 NA NA
> 33 132102 soc 6.85 19.72 26.57 NA NA
> 34 114696 soc 7.56 20.80 28.36 NA NA
> 35 114954 soc 22.29 60.28 82.57 NA NA
> 36 136206 aio 0.10 0.23 0.33 0.33 100%
> 37 165034 aio 0.10 0.23 0.33 1.10 30%
> 38 177362 aio 0.07 0.17 0.24 1.06 22%
>
> tot 6110.67 1283.52 7394.19
>
> _________________________________________________________________
> New Windows 7: Find the right PC for you.
>
> http://www.microsoft.com/windows/pc-scout/default.aspx?CBID=3Dwl&ocid=3D=
> PID24727::T:WLMTAGL:ON:WL:en-US:WWL_WIN_pcscout:112009...
>
> ***********************************************************************=
> ********
>
> Forum Note: Use "Reply" to post a response in the discussion forum.
>
> =
>
>
>
>
*******************************************************************************
> Forum Note: Use "Reply" t
Related threads
- Posting from the Informix-list
- Migrating from IDS 9.40.UC6 to 11.50.UC3
- Ip for a network session
- questions onstat -g