How to find when was my portal down

[Former Member]

Hi Experts,

             We have Netweaver Dual Stack Server. Two days ago we Restarted the Server for Weekly Cold Backup. At that time, both stacks were

started and Portal was also available. But Today we found that Portal is not running (Url is not running) . We have checked and found that

issue with " cannot attach shared memory rc =3" (using jcmon pf=<instance profile>) . To resolve this issue, we have bounced entire server

and Now it is Portal is running. My doubts are

1.  I need to know when was the portal down . i checked the "available.log" file, but it couldnt help me. Is any file/log to check that?

2.  Why do u need to restart  both stacks to resolve this issue. Because ABAP was running fine. is it needed, if yes, please clarify..

Waiting for your suggestions.

Regards,

Santosh K

[Former Member]

Hi Santhosh,

Answers to your questions:

1. There are lot of files where you can check the java start up logs. The dev_sever<n>.log

file is the best possible log file where you can find the detailed start up logs.

2. Well, from personal experience, it is the java stack which is the "bad kid" among the two. The memory error can be due to a number of reasons. Primarily some of them:

• JVM not being properly loaded - error needs to be analysed.
• Incorrect JVM Memory settings (Need to check the config tool).

For details you can refer http://scn.sap.com/message/958222#958222 or you can check the note 724452.

On the other note, The best possible solution can be found at http://scn.sap.com/thread/1516744.

Thanks,

Avishek

[Former Member]

Hi Santhosh,

1. Since "available.log" doesnt have more info, you need to check in the file std_server0.out of  /work directory for downtime info. Also see dev_server0 file for related error.

2. You can restart your JAVA Instance alone from "SAP Management Console" (SAP Management Console can be accessed as http://<servername>:5<instance number>13

Regards, Praveen

[bxiv]

This is the purpose of SolMan to be a central location to configure notifications to let you know when your systems go down unplanned.

If the system is having an issue with shared memory, chances are you have to have ABAP release a hold on the memory so that both stacks have something to use.  Have any parameters been configured recently for memory parameters?