#49000 ADD/DEL received during total update can be missing in the changelog of initilaized replica
Opened 3 years ago by lkrispen. Modified 3 months ago

If a server receives updates while performing a total update (more exactly between sending the supplier RUV and building the idlist for entries to send) these updates can be part of the database and will be sent in the total init butthey are newer than the RUV sent.

After the total init is completed the consumer side of the init set its RUV (== sent supplier RUV), then the incremental protocal starts and will send updates newer than the RUV, including the updates already sent in the init.

On the consumer side ADDs and DELs will be ignored:

[05/Oct/2016:11:45:11 +0200] conn=7 op=4 csn=57f4cc96000200640000 - urp_add (cn=xx-186,ou=People,dc=example,dc=com): an entry with this uniqueid already exists.
[05/Oct/2016:11:45:11 +0200] conn=7 op=5 csn=57f4cc98000000640000 - urp_add (cn=xx-187,ou=People,dc=example,dc=com): an entry with this uniqueid already exists.
[05/Oct/2016:11:45:12 +0200] conn=7 op=6 csn=57f4cc98000100640000 - urp_add (cn=xx-188,ou=People,dc=example,dc=com): an entry with this uniqueid already exists.
[05/Oct/2016:11:45:12 +0200] conn=7 op=7 csn=57f4cc98000200640000 - urp_add (cn=xx-189,ou=People,dc=example,dc=com): an entry with this uniqueid already exists.
[05/Oct/2016:11:57:06 +0200] conn=14 op=4 csn=57f4cf60000000640000 - urp_delete: Entry "nsuniqueid=53865a69-8ae011e6-ab8d8005-8430f734,cn=xx-179,ou=People,dc=example,dc=com" is already a Tombstone.
[05/Oct/2016:11:57:06 +0200] conn=14 op=5 csn=57f4cf60000100640000 - urp_delete: Entry "nsuniqueid=53865a67-8ae011e6-ab8d8005-8430f734,cn=xx-178,ou=People,dc=example,dc=com" is already a Tombstone.
[05/Oct/2016:11:57:06 +0200] conn=14 op=6 csn=57f4cf61000000640000 - urp_delete: Entry "nsuniqueid=53865a65-8ae011e6-ab8d8005-8430f734,cn=xx-177,ou=People,dc=example,dc=com" is already a Tombstone.

this means they will not be written to the changelog, and can be missing csns in a replication session originating from this consumer.

The probabilty for these kind of changes might be low, but we also do it systematically when creating "keep alive" entries.


Metadata Update from @lkrispen:
- Issue set to the milestone: 1.3.5.14

2 years ago

@lkrispen - Is your recent work on replication conflicts going to address this by any chance?

Metadata Update from @mreynolds:
- Issue close_status updated to: None

2 years ago

Metadata Update from @mreynolds:
- Issue set to the milestone: 1.3.6.0 (was: 1.3.5.14)

2 years ago

Metadata Update from @mreynolds:
- Issue set to the milestone: 1.3.7.0 (was: 1.3.6.0)

2 years ago

Metadata Update from @mreynolds:
- Custom field reviewstatus adjusted to None
- Issue set to the milestone: 1.4.2 (was: 1.3.7.0)

3 months ago

Login to comment on this ticket.

Metadata