VRTSexplorer Logs

[Faisal_Saleem]

Dear Experts

One of my client is running Veritas Cluster Server on Windows, Just to check how are things going on I collected VRTSexplorer logs

What I wanted to ask is just for the sake of Health Check of the enviorment which logs should I go through ?

I have checked engine log and events but they are very brief, its hard to follow them up if i wanted to view a whole month status.

Any recomendation of suggestion for easyness ?

Thanks

[mikebounds]

The engine log is a good place to start and this will tell you if anything has gone wrong, but if things have gone wrong, then it is not that good at telling you why as for instance if an application can't start, then quite often the error produced by the application is not shown in the engine log unless you set debugging levels (and sometimes even then it is not seen).  The VCS agent logs sometimes help, but they usually don't help either, so if you see an issue, you are probably best looking at the Windows Event logs (in the VRTSexplorer) to see if there is an an error message in there or any logs that the application produces.

Mike

[Satish_K__Pagar]

Have you looked at the VOM offering? It provides a very comprehensive solution for a data center. You can run various reports, checks, etc. It would need to be setp on a separate system as the VOM central server and then add your cluster nodes as the managed nodes to the VOM central server. http://www.symantec.com/veritas-operations-manager

[Faisal_Saleem]

Thanks mikebounds for your words.

I have checked the engine log and the events too, basically I wanted to look and verify a whole month logs in order to inform client how the solution is working so far and if any issue occured is already in their knowledge or not.

But I faced 2 issues

Firstly engine log for Windows is in way to much detail and print each and every change, if I look for a month history it really hard to catch up the flow

Secondly with the events they are also in a huge quantity and its hard to maintain the flow because every single day contains hundreds of event

Any other help I can have ?

[Faisal_Saleem]

Thanks satish I will look forward for it :)

[mikebounds · Accepted solution]

If there are lots of entries in the engine log then this could indicate an issue as in theory the only logs in the engine log for a cluster where the services remains online should be config changes and adminstrators logging onto the GUI - this is not always the case as some agents are poorly written and log every monitor cycle when there are no issues.  If you just filter the logs by "ERROR", then this should show you if anything went wrong.  I don't know much about the Windows event log, but I would think you could also filter this for just error logs.

I would recommend using VOM - this is a centralised management tool which must be installed on a Windows/Linux/Unix server such as Windows 2003/2008 or RedHat Enterprise Linux in your customers data centre and then the customer can access VOM via a web browser on their client desktop to operate the cluster and can set it up to receive real-time alerts and you can run reports monthly to see if there were any issues.

The VCS java GUI is been phased out so your customer may be using VOM already and if they are not, they will have to use it at some point unless they only administer cluster from the command line

Mike

[Faisal_Saleem]

Thanks mikebounds for your words,

I would look into it

Regards