Hi Rich,

 

actually just restarting srvA seems to have cleared the replication issue. It looks like replication is working fine now,

but I see now the following error log:

[18/Oct/2011:13:09:57 +0000] NSMMReplicationPlugin - agmt="cn=srvAtosrvB" (srvB:389): changelog iteration code returned a dummy entry with csn 4e9d7bc2000000080000, skipping ...

I think that we can ignore this message, right? But, how can I get rid of this message, since it is generated quite often?

Any ideas?

 

-Reinhard

---

From: Rich Megginson [mailto:rmeggins@redhat.com]
Sent: Wednesday, October 12, 2011 4:29 PM
To: Reinhard Nappert
Cc: General discussion list for the 389 Directory server project.; Marc Sauton
Subject: Re: [389-users] Replication issue

On 10/12/2011 02:16 PM, Reinhard Nappert wrote:

Good.

 

what about the different generation ID message? Is it possible that this could be caused by a re-initialize?

Yes.

 

But then, I thought a re-initialize would fix this error, if it occurs.

In this case, it should fix this problem.

 

-Reinhard

---

From: Rich Megginson [mailto:rmeggins@redhat.com]
Sent: Wednesday, October 12, 2011 4:11 PM
To: Reinhard Nappert
Cc: General discussion list for the 389 Directory server project.; Marc Sauton
Subject: Re: [389-users] Replication issue

On 10/12/2011 02:08 PM, Reinhard Nappert wrote:

Rich,

 

I was thinking about the "Replica has a different generation ID than the local data." error, because I have seen this before. If possible, I want to avoid that I have to go though each box and re-initialize.

 

So, you suggest I take let's say D (or A) and re-initialize B with D's data. Then, I would have to re-initialize F from B, right?

Right.

 

Let's go a bit further: If I had an agreement from A to F (and vice versa), I would not even have to re-initialize F from B. Is this correct?

Assuming the AtoF agreement is not complaining about "unable to find CSN" and "data reload", then yes.

 

-Reinhard

---

From: Rich Megginson [mailto:rmeggins@redhat.com]
Sent: Wednesday, October 12, 2011 4:00 PM
To: General discussion list for the 389 Directory server project.
Cc: Reinhard Nappert; Marc Sauton
Subject: Re: [389-users] Replication issue

On 10/11/2011 02:41 PM, Reinhard Nappert wrote:

How do I do this manually on server A?

 

The other question is, what kind of impact does it have when I re-iitialize server B? To be more precise, my replication environment is more complex than just server A and server B. In fact, I have a setup like the following:

 

srv C <--> srv A <--> srv B <--> srv D <--> srv C

                 /\             /\

                 |               |

                \/              \/

              srv E          srv F

 

I don't want to end up to re-initialize all boxes in my environment.

Assuming C and D are up to date and don't have any problems, reinitializing B should affect only B and F.

 

Thanks,

-Reinhard

---

From: Marc Sauton [mailto:msauton@redhat.com]
Sent: Tuesday, October 11, 2011 4:36 PM
To: General discussion list for the 389 Directory server project.
Cc: Reinhard Nappert
Subject: Re: [389-users] Replication issue

On 10/11/2011 01:22 PM, Reinhard Nappert wrote:

Hi,

 

I encountered the following logs in the errors:

[06/Oct/2011:10:11:57 +0000] NSMMReplicationPlugin - changelog program - agmt="cn=srvAtosrvB" (srvB:389): CSN 4e8d804a0000000c0000 not found, we aren't as up to date, or we purged
[06/Oct/2011:10:11:57 +0000] NSMMReplicationPlugin - agmt="cn=srvAtosrvB" (srvB:389): Data required to update replica has been purged. The replica must be reinitialized.
[06/Oct/2011:10:11:57 +0000] NSMMReplicationPlugin - agmt="cn=srvAtosrvB" (srvB:389): Incremental update failed and requires administrator action
 
Does anyone have an idea, what could have caused this and more importantly, how to fix this?
 
Thanks
-Reinhard

--
389 users mailing list
389-users@lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/389-users

On server A, read a changelog to manually run the changes on server B.
May be tune up nsds5ReplicaPurgeDelay if such errors somehow appears regularly.
Otherwise, like the errors log says, the change was purged/removed, and replica need a re-init.
M.

--
389 users mailing list
389-users@lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/389-users