Event Status : "Failed - no restart"

[rchoudarapu]

TSM Server 5.5.1

Recently, I am seeing some events show up their status as "Failed - no restart" when run a Q EVENT. And the backup sessions continue to progress.

tsm: TSM>q event * * n=SOME_NODE begind=-6 endd=today
Scheduled Start Actual Start Schedule Name Node Name Status
-------------------- -------------------- ------------- ------------- ---------
08/23/2008 19:00:00 08/23/2008 19:01:43 WI_1900 SOME_NODE Completed
08/24/2008 19:00:00 08/24/2008 19:01:37 WI_1900 SOME_NODE Severed
08/25/2008 19:00:00 08/25/2008 19:00:56 WI_1900 SOME_NODE Failed - no restart
08/26/2008 19:00:00 WI_1900 SOME_NODE Missed
08/27/2008 19:00:00 WI_1900 SOME_NODE Missed

08/28/2008 19:00:00 WI_1900 SOME_NODE Missed
08/29/2008 19:00:00 WI_1900 SOME_NODE Future

What does "Failed - no restart" mean?

TIA - RK

 

[ttrinh7975]

rchoudarapu-


This usually TSM Server is being down for a long period of time and TSM client and server lost connection. All you needs is to recycle TSM scheduler services on client side.

 

[rchoudarapu]

Failed - no restart

The events which show up the status: "Failed - no restart" are not reporting any failed backup stats in actlog and actually the backup is progressing after the event status is updated as Failed - no restart.

Policy Domain Name: XX_DOMAIN
Schedule Name: WI_2100
Node Name: NODE_NAME
Scheduled Start: 09/08/2008 21:00:00
Actual Start: 09/08/2008 21:00:25
Completed: 09/09/2008 07:51:16
Status: Failed - no restart
Result:
Reason:

And the actlog reports the following at 09/09/2008 07:51:16:

09/09/2008 07:51:16 ANR0481W Session 79802 for node NODE_NAME (WinNT) terminated - client did not respond within 3600 seconds.

 

[nnss]

TSM Server 5.5.1

tsm: TSM>q event * * n=SOME_NODE begind=-6 endd=today
Scheduled Start Actual Start Schedule Name Node Name Status
-------------------- -------------------- ------------- ------------- ---------
08/23/2008 19:00:00 08/23/2008 19:01:43 WI_1900 SOME_NODE Completed
08/24/2008 19:00:00 08/24/2008 19:01:37 WI_1900 SOME_NODE Severed
08/25/2008 19:00:00 08/25/2008 19:00:56 WI_1900 SOME_NODE Failed - no restart
08/26/2008 19:00:00 WI_1900 SOME_NODE Missed
08/27/2008 19:00:00 WI_1900 SOME_NODE Missed

08/28/2008 19:00:00 WI_1900 SOME_NODE Missed
08/29/2008 19:00:00 WI_1900 SOME_NODE Future

What does "Failed - no restart" mean?

TIA - RK

Hi, I had a similar problem with an AIX server that at certain time, all the FSs ar dismounted to do a flashcopy (IBM shark to other one). An the dsmc sched is killed when is backing up some files of the directory that are dismounted. And in the servers log I see the "serverd" status.

is anything wrong in your agent host (a kind of warning in dsmerror file)?

 

[Lars_Svensson]

Same thing here. A lot schedules reports as Failed - no restart when running a Q EVENT.
These sessions has timed out and then restarted and the backup usually completes successfully.

TSM Server 5.5.1 WIN 2003

 

[aroon_e]

Hi,

I am facing the same problem daily in two of our windows servers.But the report says backup completed successfully.

TSM Server 5.5.0.3 Windows 2003.

Aroon

 

[ChrisRees]

Hi

I've got one node that everyday shows Failed - no restart..
This backup does not complete and I have to restart the scheduler everyday.

There are no errors in dsmerror.log.

I've just raised a call with IBM about it.
This is a 5.5.1.6 windows 2003 client talking to a 5.5.1 AIX TSM Server.

Cheers

 

[fakesugar]

Got same problem on 2 client nodes. Have to restart the TSM schedule services.

However, ran the one time backup for them during the working hours and the backup was done successfully. Looks weird.

 

[DFrance]

These are typical anomalies of the "q event" results... there is a long history of distorted results posted in the events table; different release-levels have varying degrees of accuracy!

This is why I recommend customers install and customize Servergraph (best-of-breed) or (at least) operational reports (with appropriate "tweaking" to compensate for nodes that consistently get incorrect, false failure or success reports).

Servergraph resolves the scheduled events issue (at least does a better job than the "q event" tables) using actual behavior of client and server actions based on activity log (and other table updates).

HTH!

-Don

 

[PJ]

"Servergraph (best-of-breed)"

Very funny...

If your clients don't respond within the commtimeout/idletimeout windows but are in fact working, you may want to consider increasing these parameters or tune the clients appropriately (resourceutil etc.) or find out what else may lead to such a condition - network or routing problems come to mind.

PJ