Hi, OK, I have a problem with MultiMaster Replication between my 3 x 1.01-4 FDS. I have already tried the following: Stop servers A and C Delete the contents of the changelog database directory restart push a replication to each from Server B (successful) While my data was momentarily in sync, I immediately went back to the problem where incremental updates (users updating passwords) failed. Here is a typical set of errors from logs/errors on server A: [07/Mar/2008:15:46:31 -0500] agmt="cn="Replication to B.company.com"" (B:389) - Can''t locate CSN 47d04e2e000000010000 in the changelog (DB rc=-30990). The consumer may need to be reinitialized. [07/Mar/2008:15:46:31 -0500] agmt="cn="Replication to C.company.com"" (C:389) - Can''t locate CSN 47d04e2e000000010000 in the changelog (DB rc=-30990). The consumer may need to be reinitialized. The logs on B and C are analogous. Both have the following (This from C): [07/Mar/2008:07:46:48 -0500] - import userRoot: Workers finished; cleaning up... [07/Mar/2008:07:46:48 -0500] - import userRoot: Workers cleaned up. [07/Mar/2008:07:46:48 -0500] - import userRoot: Indexing complete. Post-processing... [07/Mar/2008:07:46:48 -0500] - import userRoot: Flushing caches... [07/Mar/2008:07:46:48 -0500] - import userRoot: Closing files... [07/Mar/2008:07:46:49 -0500] - import userRoot: Import complete. Processed 496 entries in 5 seconds. (99.20 entries/sec) [07/Mar/2008:07:46:49 -0500] NSMMReplicationPlugin - multimaster_be_state_change: replica dc=company,dc=com is coming online; enabling replication [07/Mar/2008:07:46:49 -0500] NSMMReplicationPlugin - replica_reload_ruv: Warning: new data for replica dc=company,dc=com does not match the data in the changelog. Recreating the changelog file. This could affect replication with replica''s consumers in which case the consumers should be reinitialized. The user data started out on 7.1 on Directory Server "A". I installed and replicated it to servers B and C and ran it like that for a while, and replication worked fine. Then I replaced 7.1 with 1.01-4 on A (just as 1.1 was released), and replication has been broken since. I am willing to lose whatever updates or differences necessary to get these sync''ed up ASAP, any tips? Thanks for looking, Ken Marsh
Rich Megginson
2008-Mar-07 21:22 UTC
Re: [Fedora-directory-users] 3-way MMR problem on 1.01-4
Ken Marsh wrote:> Hi, > > OK, I have a problem with MultiMaster Replication between my 3 x 1.01-4 > FDS. I have already tried the following: > > Stop servers A and C > Delete the contents of the changelog database directory > restart > push a replication to each from Server B (successful) > > While my data was momentarily in sync, I immediately went back to the > problem where incremental updates (users updating passwords) failed. > > Here is a typical set of errors from logs/errors on server A: > > [07/Mar/2008:15:46:31 -0500] agmt="cn="Replication to B.company.com"" > (B:389) - Can''t locate CSN 47d04e2e000000010000 in the changelog (DB > rc=-30990). The consumer may need to be reinitialized. > [07/Mar/2008:15:46:31 -0500] agmt="cn="Replication to C.company.com"" > (C:389) - Can''t locate CSN 47d04e2e000000010000 in the changelog (DB > rc=-30990). The consumer may need to be reinitialized. > > The logs on B and C are analogous. Both have the following (This from > C): > > [07/Mar/2008:07:46:48 -0500] - import userRoot: Workers finished; > cleaning up... > [07/Mar/2008:07:46:48 -0500] - import userRoot: Workers cleaned up. > [07/Mar/2008:07:46:48 -0500] - import userRoot: Indexing complete. > Post-processing... > [07/Mar/2008:07:46:48 -0500] - import userRoot: Flushing caches... > [07/Mar/2008:07:46:48 -0500] - import userRoot: Closing files... > [07/Mar/2008:07:46:49 -0500] - import userRoot: Import complete. > Processed 496 entries in 5 seconds. (99.20 entries/sec) > [07/Mar/2008:07:46:49 -0500] NSMMReplicationPlugin - > multimaster_be_state_change: replica dc=company,dc=com is coming online; > enabling replication > [07/Mar/2008:07:46:49 -0500] NSMMReplicationPlugin - replica_reload_ruv: > Warning: new data for replica dc=company,dc=com does not match the data > in the changelog. > Recreating the changelog file. This could affect replication with > replica''s consumers in which case the consumers should be > reinitialized. > > The user data started out on 7.1 on Directory Server "A". I installed > and replicated it to servers B and C and ran it like that for a while, > and replication worked fine. Then I replaced 7.1 with 1.01-4 on A (just > as 1.1 was released), and replication has been broken since. > > I am willing to lose whatever updates or differences necessary to get > these sync''ed up ASAP, any tips? >Looks like you may be running into https://bugzilla.redhat.com/show_bug.cgi?id=388021 which was fixed in Fedora DS 1.1. One of the comments has a workaround from another Fedora DS 1.0.x user.> Thanks for looking, > Ken Marsh > > > -- > Fedora-directory-users mailing list > Fedora-directory-users@redhat.com > https://www.redhat.com/mailman/listinfo/fedora-directory-users >