Hi,
It seems weird results are popping up faster than I can assemble
test-setups to reproduce.
I ran a test in mirrormode were I:
1) Took an slapcat generated LDIF from a 2.3.x setup
2) Removed all entryCSN and contextCSN lines.
3) Ran slapadd -S 1 -q -w -l ~/load_noCSN.ldif on server-1
4) Did a slapcat toserver2.ldif on server-1
5) Started server-1 and let applications create and modify objects.
6) Moved toserver2.ldif to server-2.
7) Ran slapadd -q -l toserver2.ldif on server-2
8) Started server-2
Now - I would expect the objects created on step 5 to appear after a
while on server-2. They are not.
The reason seems to be that the contextCSN has not been updated properly
on server-1.
The contextCSN in the toserver2.ldif file from step 4 is:
20091118173357.690685Z#00#001#00
The contextCSN on both servers are now:
20091119140605.714382Z#00#001#00
The entry on server-1 having that entryCSN is a modified object,
however, the change has not been replicated to server-2.
Likewise, the entryCSN for one created object in step 5 is:
20091119094729.719193Z#00#001#00
It is not present on server-2, but it appears TWICE in a the result from
an ldapsearch (scope sub) on its parent object on server-1.
?!?! .. I'm lost.
From all I know this should not happen on a healthy setup - unless
there's something badly wrong with the procedure I've described above.
It's slapd 2.4.19 with BDB4.8
/Peter