Production server Databse & Application Server hang

[Former Member]

Dear ALlt ,

Yesterday , After whole Offline + redolog Backup finish successfully,

my Database and Application Servers becomes hang. ( Nobody User logn to

the SAP ).

I able to Login telnet , to see the Lnrctl status , is hang.

so ,using sqlpls shut down oracle nortmal, restart the server.

After the Reboot of Database and Application Server System works

Normal.Now I am unable to find out Why the Server becomes hang ?

So avoid this Error in future , Please suggest the solution.

We are running SAP Ecc 6.0 + Oracle 10.2.0.2 + Hp unix 11.23

Thanks is advance

Santosh

[former_member204746 · Accepted solution]

increase parameters SESSIONS and PROCESSES as per SAP note 830576.

[former_member184473]

Hi Santosh,

You can also check note [521264|https://service.sap.com/sap/support/notes/521264] for further information.

Note 521264 - Hang situations

Regards,

Eduardo Rezende

[Former Member]

Hi

pls Upgrade your Oracle patch set from 10.2.0.2 to 10.2.0.4 and interm patches also ...

[Former Member]

Hi ,

Please refer to note 1414417 for further information

[Former Member]

Hi ,

It is highly recommended to apply 10.2.0.4 as a lot of other problems

are as well solved within 10.2.0.4 that might be encountered with Oracle10.2.0.2.

If you want to apply 10.2.0.4, please follow notes #871735 and #839187.

They explain how to download and apply 10.2.0.4. Please note that you

also need to apply the most recent interims patches on to of 10.2.0.4.

Note #1137346 explains how to download and apply the recent PSU bundle

which contains all required interims patches.

[volker_borowski2]

Hi,

I have seen such behavior only on windows systems that

have been rarely booted and used offline Backup via DB13

while SAP keeps running. After a while

lsnrctl services

may show an excessive number of comslots in use.

It is new for me that this can happen on Unix as well.

I have customers runnning one offline backup a week, and

in general, on W2K3 you'll get along for one and a half year.

So when you are tight on security fixes, in this amount of

time you generally have a reboot, so no problem occurs.

(But one of my customers is very lazy in this area, we had

two of these hangs at his site. Was with 9.2 and 10.2 of Oracle.)

Btw: on windows it does not help to simply restart the listener.

Allthough the amount of sevice connection drops instantly.

I think it has to be related with the re-connect retries SAP is doing

in offline backup, because system only doing online backups show

a lot less of these used slots.

Did not find anything to fix this since yet. And no, we did not log any

call for this, because noboy wants to wait for support reaction, when

a boot fixes it and it is comming only every 18 months.

Volker