Author |
Message
|
trystan2k |
Posted: Tue Oct 23, 2012 3:00 pm Post subject: |
|
|
Novice
Joined: 22 Oct 2012 Posts: 22
|
Just to clarify, machine on PROD is more powerful (not much but more memory, cpu) than DEV, which does not make much sense that prod has worse performance than DEV (I expect the opposite) |
|
Back to top |
|
 |
ramires |
Posted: Tue Oct 23, 2012 3:42 pm Post subject: |
|
|
Knight
Joined: 24 Jun 2001 Posts: 523 Location: Portugal - Lisboa
|
I tried you script in one box with MQ 7.5. Run five times and the result was
18; 18; 18; 18;19 secs.
Maybe the PROD environment has more work to do (other applications)?
Better to talk with your MQ admin team _________________ Obrigado / Thanks you |
|
Back to top |
|
 |
trystan2k |
Posted: Tue Oct 23, 2012 4:06 pm Post subject: |
|
|
Novice
Joined: 22 Oct 2012 Posts: 22
|
ramires wrote: |
I tried you script in one box with MQ 7.5. Run five times and the result was
18; 18; 18; 18;19 secs.
Maybe the PROD environment has more work to do (other applications)?
Better to talk with your MQ admin team |
Actually no, again for our surprise, the PROD machine is not live yet, so there is no traffic at all while DEV there are some small...
Seems to be such a small setting or tweak missing but we cannot find it! |
|
Back to top |
|
 |
ramires |
Posted: Tue Oct 23, 2012 4:17 pm Post subject: |
|
|
Knight
Joined: 24 Jun 2001 Posts: 523 Location: Portugal - Lisboa
|
Did you check MQ error logs ? Are there any FDC files created (in the PROD box) |
|
Back to top |
|
 |
trystan2k |
Posted: Wed Oct 24, 2012 5:28 am Post subject: |
|
|
Novice
Joined: 22 Oct 2012 Posts: 22
|
ramires wrote: |
Did you check MQ error logs ? Are there any FDC files created (in the PROD box) |
Hi, logs does not say much, it has more about the when I tried to connect using 'localhost':
Code: |
The attempt to allocate a conversation using TCP/IP to host 'localhost
(127.0.0.1) (9494)' for channel EAI.CLIENT.SVRCONN was not successful. However
the error may be a transitory one and it may be possible to successfully
allocate a TCP/IP conversation later.
In some cases the remote host cannot be determined and so is shown as '????'.
ACTION:
Try the connection again later. If the failure persists, record the error
values and contact your systems administrator. The return code from TCP/IP is
111 (X'6F'). The reason for the failure may be that this host cannot reach the
destination host. It may also be possible that the listening program at host
'localhost (127.0.0.1) (9494)' was not running. If this is the case, perform
the relevant operations to start the TCP/IP listening program, and try again. |
Also there is 5 FDC files, but none are from yesterday tests... I don´t know how to 'read' this FDC files? I mean, where I need to search for useful information ? Is there an app that open it to better read?
Thanks ! |
|
Back to top |
|
 |
ramires |
Posted: Wed Oct 24, 2012 5:38 am Post subject: |
|
|
Knight
Joined: 24 Jun 2001 Posts: 523 Location: Portugal - Lisboa
|
The FDC is a text file , with a readable header.
Quote: |
(127.0.0.1) (9494)' |
This message shows tcp port 9494 being used , but in you test you have port 1414. Why is this? _________________ Obrigado / Thanks you |
|
Back to top |
|
 |
trystan2k |
Posted: Wed Oct 24, 2012 5:52 am Post subject: |
|
|
Novice
Joined: 22 Oct 2012 Posts: 22
|
ramires wrote: |
The FDC is a text file , with a readable header.
Quote: |
(127.0.0.1) (9494)' |
This message shows tcp port 9494 being used , but in you test you have port 1414. Why is this? |
I have two Queue Managers... One running at 1414 and other at 9494... I have the same error entry for 1414... |
|
Back to top |
|
 |
trystan2k |
Posted: Wed Oct 24, 2012 5:57 am Post subject: |
|
|
Novice
Joined: 22 Oct 2012 Posts: 22
|
ramires wrote: |
The FDC is a text file , with a readable header.
Quote: |
(127.0.0.1) (9494)' |
This message shows tcp port 9494 being used , but in you test you have port 1414. Why is this? |
OK, just did some research and found that this could be some interessing fields in FDC:
Code: |
Probe Id :- RM557001 |
| Application Name :- MQM |
| Component :- rriAsyncConvControl |
| SCCS Info :- lib/remote/amqrmasa.c, 1.97 |
| Line Number :- 1259 |
| Build Date :- Oct 14 2011 |
| CMVC level :- p000-L111015 |
| Build Type :- IKAP - (Production) |
| Effective UserID :- 510 (mqm) |
| Real UserID :- 501 (thiago) |
| Program Name :- amqrmppa |
| Addressing mode :- 64-bit |
| LANG :- en_US.UTF-8 |
| Process :- 11099 |
| Process(Thread) :- 3558 |
| Thread :- 1043 |
| ThreadingModel :- PosixThreads |
| QueueManager :- PRODQM |
| UserApp :- FALSE |
| ConnId(1) IPCC :- 7241 |
| ConnId(3) QM-P :- 4087 |
| Last HQC :- 3.0.0-40688 |
| Last HSHMEMB :- 0.0.0-0 |
| Major Errorcode :- rrcE_PROTOCOL_ERROR |
| Minor Errorcode :- OK |
| Probe Type :- MSGAMQ9504 |
| Probe Severity :- 2 |
| Probe Description :- AMQ9504: A protocol error was detected for channel ''. |
|
But as i said, this FDC are from 3 days ago... And I did a lot of tests yesterday and no FDC was generated by that date...
Last edited by trystan2k on Wed Oct 24, 2012 6:04 am; edited 1 time in total |
|
Back to top |
|
 |
ramires |
Posted: Wed Oct 24, 2012 6:04 am Post subject: |
|
|
Knight
Joined: 24 Jun 2001 Posts: 523 Location: Portugal - Lisboa
|
|
Back to top |
|
 |
trystan2k |
Posted: Wed Oct 24, 2012 6:06 am Post subject: |
|
|
Novice
Joined: 22 Oct 2012 Posts: 22
|
Ok, thanks for the link ! I will read it...
But as I said, it not seems that this FDC would have anything with my tests (otherwise others would be created yesterday while I was testing right ?)..
Thanks ! |
|
Back to top |
|
 |
ramires |
Posted: Wed Oct 24, 2012 6:11 am Post subject: |
|
|
Knight
Joined: 24 Jun 2001 Posts: 523 Location: Portugal - Lisboa
|
One FDC file can have more than one entry. You can do a backup for the existing FDC files , clean the "errors" directory and test again. |
|
Back to top |
|
 |
ramires |
Posted: Wed Oct 24, 2012 6:16 am Post subject: |
|
|
Knight
Joined: 24 Jun 2001 Posts: 523 Location: Portugal - Lisboa
|
|
Back to top |
|
 |
trystan2k |
Posted: Wed Oct 24, 2012 6:30 am Post subject: |
|
|
Novice
Joined: 22 Oct 2012 Posts: 22
|
Checked, and all are set to more values than the minimum, except for max file...
Also comparing each other, there are only one values that are higher in DEV than in PROD, shmall (The maximum amount of shared memory that can be allocated).
Do you think that these 2 values (file-max and shmall) could be related to the slow performance in connecting to the MQ ?
Not sure about file-max, as in DEV it runs fine with this value (and in PROD we do not have anything running that could have open files, as this box is not live yet)
Code: |
PROD
$ cat /proc/sys/kernel/msgmni = 196608
$ cat /proc/sys/kernel/shmmni = 8192
$ cat /proc/sys/kernel/shmall = 102760448
$ cat /proc/sys/kernel/shmmax = 206158430208
$ cat /proc/sys/kernel/sem = 500 25088000 282 50175
$ cat /proc/sys/fs/file-max = 200000
$ cat /proc/sys/net/ipv4/tcp_keepalive_time = 300
DEV
$ cat /proc/sys/kernel/msgmni = 6144
$ cat /proc/sys/kernel/shmmni = 7168
$ cat /proc/sys/kernel/shmall = 4294967296
$ cat /proc/sys/kernel/shmmax = 68719476736
$ cat /proc/sys/kernel/sem = 500 716800 250 7168
$ cat /proc/sys/fs/file-max = 200000
$ cat /proc/sys/net/ipv4/tcp_keepalive_time = 300
|
|
|
Back to top |
|
 |
ramires |
Posted: Wed Oct 24, 2012 6:41 am Post subject: |
|
|
Knight
Joined: 24 Jun 2001 Posts: 523 Location: Portugal - Lisboa
|
why don't you try as recommended in the documentation (Infocenter) ?
Code: |
"The minimum configuration for WebSphere MQ for these resources is as follows:
kernel.msgmni = 1024
kernel.shmmni = 4096
kernel.shmall = 2097152
kernel.shmmax = 268435456
kernel.sem = 500 256000 250 1024
fs.file-max = 524288"
|
|
|
Back to top |
|
 |
trystan2k |
Posted: Wed Oct 24, 2012 6:47 am Post subject: |
|
|
Novice
Joined: 22 Oct 2012 Posts: 22
|
ramires wrote: |
why don't you try as recommended in the documentation (Infocenter) ?
Code: |
"The minimum configuration for WebSphere MQ for these resources is as follows:
kernel.msgmni = 1024
kernel.shmmni = 4096
kernel.shmall = 2097152
kernel.shmmax = 268435456
kernel.sem = 500 256000 250 1024
fs.file-max = 524288"
|
|
I will try to increase the file-max, as I think this is the only one that is less then min, and could have something to do. All others are higher than the min which I guess give more 'power' to the environment...
Thanks |
|
Back to top |
|
 |
|