Shortcuts still pending after Centera replication

[cwapshere]

We add an item to the archive. We can see it stage through the Collections folder. We can see the number of files increase by 1 on Cenetra1, and then subsequently on Centera2. But the shortcuts remain in a "Pending" state.

What information does EV need back from the Centera that it is not getting?

We have EV 8.0 sp1

Thanks,
Carol

[Liam_Finn1]

For Centera you need to set the vault store properties to Remove safety copies immediately after archive

[cwapshere]

Thanks for your response Liam.

The EV documentation does not seem to agree with you. From "Enterprise Vault Administration Console" / "New vault store":

<quote>If you intend to use a Centera device for the new vault store, "After backup" really means "After replication", so if you use Centera replication, select After backup.<\quote>


That certainly seems to be telling me to select "After backup"... No?


Carol
 

[MichelZ]

Carol

You are correct, if you have replicating Centeras, you should be fine.
Have you added both of your Centeras to the PEA file?
Are you able to access both with CenteraVerify.exe?

Cheers

[Liam_Finn1]

We have two Centera devices with one way replication. Symantec support told us to use immediate after archive for the configuration settings

[MichelZ]

Liam

I have successfully used it before with "After Backup", it has to work...
But you need to have the 2nd Centera setup as Replication Centera in EV.

/Michel

[cwapshere]

CenteraVerify is successful for both of them. What is not successful is the replication test. Next to "Replicate" I see "No primary cluster found". I thought that perhaps this is because the replication has been configured to go both ways. The replication is certainly working.

Carol

[MichelZ]

So, that's most likely the problem, that EV does not recognize the replication.
I'm not a centera guru, you should speak to your EMC folks for this, I think.
Seems like two-way-replication is not supported?

[TonySterling]

any event id 6760 in the app log?

This sounds similar to your issue.

http://seer.entsupport.symantec.com/docs/314540.htm

[cwapshere]

No errors like that.

Why is it talking about a "default pool"? We created a new pool for EV archiving.

Thanks,
Carol

[TonySterling]

You may  want to verify that.

show pool detail default

[SolarP]

Hi Carol

See my post below from last year it gives a few commands to run on the Centera and also an explanation of what EV is looking for when it campare the replicated clips as I had a similar problem on a new install

http://www.symantec.com/connect/forums/centra-and-safety-copies-best-option

Hope it helps

Julian

[MichelZ]

Any news here?

Cheers

[susika]

any news ?

2 centeras on 2 locations. Replication is working, but EV isn't able to show "replica ip adress" in the mailboxstore. Field is empty. The connection is availabe via NAT, centera ping and centera verify are fine. Even EVCentera Checker shows savesets in the store. Shortcuts still pending.
Maybe EV isn't supporting NAT ?

[JesusWept3]

CenteraVerify fails on the replication is due to the version of CV that is distributed with EV.
If you get a later version, then you will be able to see it go succesfully.

However in your situation, if the Centera Connections tab shows NO replica Centera ip's then this is a centera configuration problem, most likely with the user or  the inability to read the replica settings.

When EV first makes the connection to the pool, it will automatically get the replica IP's from the centera, and EV will populate it throug that call, which is why you can't edit it yourself. This is also why you have pending items, because EV isn't even looking for replication, as you would normally see in a dtrace it requesting ClipExists against the replica every few seconds.

If you call support, there is a regkey they can put in to get past this for the moment.
But you are either going to have to either change to immediately after archive or contact EMC and find out why the Replica IP address isn't reported to EV

[susika]

we found out that there are still some DNS problems with resolving adresses on 3 of 6 nodes ... when we finished with these errors, maybe it will work ....we even thought about a not matching default retention class on centera with our default on EV. I will report later on.

[susika · Accepted solution]

DNS problems solved, all nodes are accessible. We had to add two connection to the replica nodes on our firewall. Now those two replica nodes are showing up.  We had to contact the EMC support to help us. All recent  items waiting to be indexed/backed up on this centera partition are gone. So I think it works!