Java hangs in XI 3.0

[Former Member]

Hello!

There is a problem in my XI 3.0 system: 1 time in 1-2 weeks our XI server hangs. The symptoms are:

1) There are many messages in queue in transaction smq2.

2) I cannot enter rwb. I try to open it but it hangs.

3) GUI is working correctly

4) All messages are correct and pass after Java part of XI server restart.

5) All messages are not big: maximum 3 Mb.

Have you any idea?

[Former Member]

I find some messages every day which have status 'scheduled' for a long time - more then several hours. May be when have many of such messages we lose memory. These messages use http adapter. When http server is not working, the messages scheduled, but not executed after http server become working. How to execute the messages automatically?

[former_member186158]

Register the queue and activate the queue in t-code sxmb_adm.

Maybe big message or memory is not enough.

[Former Member]

Hello Shen Peng.

I check messages sizes using report RSXMB_REMOTE_SERVICE. Max message size is 1,5 Mb. I don't think it's big.

The memory is also enough.

[Former Member]

This sometimes also happen because of the Java version you are using!!! I think you are using the latest java version in your system, or some parts of Java components are been upgraded.

Check whether you are on Version 1.4

Regards,

Sainath Chutke

[Former Member]

Hello Sainath Chutke.

We have version 1.4.

[Former Member]

this is quite normal when u complete the upgrade or apply support pack new installation.

1. check out all the user are unlocked.

2, check all the background jobs (required to restart the message) are rescheduled after the upgrade or support. most of the time it will be in canceled status.

3. Is ur system clustered environment; make sure all the load balancing / web dispatcher are up and running.

4. check out ur server status.

5. check for your DB memory.

6. get the log from SM21.

7. check out for ST22 dump?

[Former Member]

Hello Prabhu Rajesh Janardanan.

The problem repeats every 1-2 weeks. We don't upgrade every week.

I followed your advice:

1) I checked users. All are unlocked.

2) All background jobs works correctly and don't have canceled status. May be, some jobs that must be planned not planned at all. What jobs must be planned?

3) I don't no how to check it.

4) Where to check it?

5) What do you mean? Free space?

6) IN SM21 there are no critical errors.

7) We have no dumps during long period.

[Former Member]

Hi,

This might happen due to insufficient system threads.

Ref: SAP NOTE 1016241 - J2EE Engine hangs because of unsufficient system threads

SAP note: 1259465: Memory analyser.

SAP note: 723909 -> set correct VM parameters

As far as I know, there are many issues with XI 3.0 SP 19. I recommend to upgrade.

Thanks,

[Former Member]

Hi, I am System Admin at this system.

This is not so correctly. Not J2ee engine hangs. Java Applications hangs. J2ee Server is working at this time, but is not possible to get RWB of XI.

Memory and treaads is correct in this system. You can see in NWA (Netweaver Administrator) - u2013 Monitoring u2013 Java System Reports.

The situation is more looks like "big message in XI". How to check it?

MichaelP

[Former Member]

Hi Michael,

If the message is huge, then the message will lie for a long time in the queue. Check it.

I think there might be some issue in the configuration or J2ee engine is getting burden with the threads. I think it is better to do a XI rediness check.

Follow SAP Note 817920 - XI Readiness Check 3.0/7.0

Thanks,