cancel
Showing results for 
Search instead for 
Did you mean: 

Failed providers in VEA

jameseaton
Level 2
Hello,

We are having  a problem with some failed storage providers in VEA stopping our cluster failover working correctly.

The 2 failed providers are VDSPROV and DDLPROV and are shown by going here in VEA - Load VEA | Localhost | Properties | choose StorageAgent and scroll bottom pane

Is there any way to reinstall the failed providers rather than reinstall the entire suite?

Thanks

james



imagebrowser image

1 ACCEPTED SOLUTION

Accepted Solutions

Jay_Kim
Level 5
Employee Accredited Certified

You can have a look at vxisis.log (not vm_vxisis.log) as it logs the load events for most (if not all) providers one by one.
It's located @ C:\Program Files\Veritas\Veritas Volume Manager 5.1\logs

Some common causes are PATH environment variable for Windows.
Windows 2003 SP1 supports up to 1023 characters. If it's over this, it may truncate it. (Not always at the end it seems)
Windows 2003 SP2 supports 2047 characters but I have seen some SP2-upgraded systems that do not still.

Other than this, you can definitely try running a Repair on the SFW server components via Add/Remove programs.
It will require a reboot in most cases and there's no guarantee but I have seen it does fix it sometimes. (quite rare however)
 

View solution in original post

3 REPLIES 3

Jay_Kim
Level 5
Employee Accredited Certified

You can have a look at vxisis.log (not vm_vxisis.log) as it logs the load events for most (if not all) providers one by one.
It's located @ C:\Program Files\Veritas\Veritas Volume Manager 5.1\logs

Some common causes are PATH environment variable for Windows.
Windows 2003 SP1 supports up to 1023 characters. If it's over this, it may truncate it. (Not always at the end it seems)
Windows 2003 SP2 supports 2047 characters but I have seen some SP2-upgraded systems that do not still.

Other than this, you can definitely try running a Repair on the SFW server components via Add/Remove programs.
It will require a reboot in most cases and there's no guarantee but I have seen it does fix it sometimes. (quite rare however)
 

Marianne
Level 6
Partner    VIP    Accredited Certified
Have a look and see if you have something similar in Event Viewer:

ERROR 1000(0x000003e8) Application Error <hostname> Faulting application vxpal.exe, version 3.3.1048.0, time stamp 0x48842240, faulting module mpioprov.dll, version 5.1.0.398, time stamp 0x48847a3a, exception code 0xc0000005, fault offset 0x00013d43, process id 0xd74, application start time 0x01c94fd587522183.

It is the same problem documented in this TechNote: http://seer.entsupport.symantec.com/docs/309219.htm

Symantec has an engineering fix that they will provide when you log a call.

rhanley
Level 4
Adding to what Jay Kim mentioned, we see path statements causing this issue quite often. Even if the entire path statement is below the numbers provided by Jay Kim, a timeout can occur at the provider level because it takes too long to traverse the path statement and find the proper directory.

If your path statement is somewhat long, I would suggest moving the Veritas paths closer to the beginning of the path and rebooting the server. The easiest way to do this is to go into the Environment Variables (My Computers > Properties > Advanced > Environment Variables), locate and open the 'Path' System variable and copy the entire line and paste it into a text editor (i.e. notepad). Once in there, move the Veritas paths closer to the beginning, then copy the modified path statement and paste it back into the Path variable entry.

Once done, reboot and see if the providers load properly.

Lastly, if this server is running Dell OPen Array Manager and changes have been made recently, this also can cause providers to fail. In those cases, SFW needs to be reinstalled after changes have been made to Dell Open Array Manager.

Hope this helps,
rhanley