Question on how to install Journal on Microsoft Cluster

[chengdj]

I followed the instruction on NOV 7 web presentation to configure the journal service on a microsoft cluster. It looks like the journal is working, because, I don't get any error messages in jberror.log, dsmerror.log, or system logs. Journal service started without any error; however, the journal database is not increasing. It is always 1 KB. I am sure that it is not working properly. Does anyone has the same experience as I do ? Thanks for all help you can give...



:sad:

 

[chengdj]

Got it to work...

 

[stroke4]

How did you get it to work???



How many objects are on the server?

How much storage is on the server?

Does the Journal service stay up or does it restart?

Is the Journal database getting any corruption errors???



Inquiring minds want to know.



Thanks.

Paul

 

[chengdj]

There are about 230,000 objects and about 40 GB. The database get about 850 KB. The Journal service does restart between fail-over and database closed and reporn without any error. I will place it on a server with 15 Millions objects and about 400 GB. we will see if this will work ...



How did I get it to work ? I follow the Instruction from Marc Lanteigne with minor correction..

 

[stroke4]

Ah HAH!!!



http://www-1.ibm.com/support/docview.wss?uid=swg21220848&aid=1



Minor correction??? Hmmm???

 

[chengdj]

When you create the directory for the journal for the cluster drive, the tech docum indicated to move those 4 files into that directory. I have not able to start the journal service. It keeps on crashing. You have to copy all the files into that directory, then the journal service will work. dbghelp, dsc30enu, dscenu, dschenu.cnt, dschenu.GID, tsmjbbd.exe, tsmjbbd.ini, tsmjbbmr.dll.

 

[itdrew]

Hi chengdj. How did the install go with a cluster that has 15 million objects?



I've installed it on hi impact Windows clusters with varied success.

 

[chengdj]

It has been running very well on both node, one exception is tsmjb_database is not resetting every time backup runs. It is running on 5.3.3 version of tsm client. Backup time has dropping from 13 hrs to 15 min per node. I still don't know why this database is not resetting itself.

 

[SkyMover]

Could someone repost the instructions on getting journalling to work in a cluster please.