Media Server not communicating with EMM but...

[owl2201]

Ok, this is a bit of a strange one, just wondering if anyone has come across this before?

We made a change to 3 of our media servers, just adding a static route to a Data Domain subnet and adding the DD into each Media Servers hosts file. Since then backups have continued to work through the media servers, but when on the media server it says it has no communication with EMM. When running a tpconfig -dsh I get the below error...

Failed to retrieve OpenStorage authentication credentials.
Unable to connect to the EMM server

All usual communication tests I have run on both media server and master server side and all come back as expected. The master sees the media servers as up and all present and correct, and backups are even running and completing! How strange. Now i know this must have something to do with the changes we made but i'm struggling to see how or why.

Anyone seen this before or have any ideas, as I'm stumped. Many Thanks.

[Mark_Solutions]

This doesnt ring any bells to be honest but i wonder if the media servers ost-plugins directory has its configuration file with something relating to the IP address of the DD

If so it may be worth deleting it (better still just move it else where for now) and then re-add the credentials for the DD to the media server to force it to be re-created properly

My gut feeling is that it is a network issue with the new routes disrupting comms to the Master but if backups actually work then maybe not

Work a look through the ost-plugins to see what is in there - there should be a .cfg file named after the storage server.

Open it in notepad and see if there is something wrong in it

[Mark_Solutions]

Worth a look here too : http://www.symantec.com/docs/TECH202796

useful checks, especially bpstsinfo -pi

[billy_no_mates]

Greets owl2201,

Well, IMHO, when it comes to anything relating to updating networks or host files in Netbackup I'd always tread with caution...perhaps 1 or 2 incidents that left me a little 'stumped' too in the past ;)

Working is good - so I'd say wait an hour, perhaps longer, and see whether it clears itself up. I came across an incident relating to updates to host files and things and waded into application caching @ NBU 7.0:

http://www.symantec.com/business/support/index?page=content&id=TECH136792

We're assuming (a filthy word!) that the backups are working ... but are they working through the correct interfaces/static routes you've added? What was the change meant to achieve? Popping DD and Media Servers all on same subnet? When you perform traceroute from each Media Server are the routes correctly going out the interface you intend and arriving at DD on correctly resolved host name? NBU thinks not ... go back to the beginning and recheck the subnets/resolution slowly. What does DD GUI say about interfaces (if you're using multiple etc etc)?

Any 'whitespace' in the host files? Probably wouldn't explain why its 'working' ... however check it again.

Good luck,

Billy.

[revarooo]

Yeah, if you're at 7.0.1 or later you can clear up the host cache by running:

bpclntcmd -clear_host_cache