[Date Prev][Date Next] [Chronological] [Thread] [Top]

Problem upgrading 2.2 -> 2.3: slap_global_control: unrecognized control

Hello Openldap,

I currently have an openldap configuration with a master server running opendlap 2.2.29
and slaves running 2.2.23, 2.2.24, 2.2.28, and 2.2.29 (5 replicas total on systems under
various administrative controls).

I'm trying to upgrade the master from 2.2.29 to the latest 2.3.x version (working
with the FreeBSD port 2.3.21). Not long ago we started running the syncrepl replication
protocol with the refresh only option, but we've found our 2.2.29 master is unstable with
that configuration (it occasionally hangs in an apparent tight loop - something we
didn't see with our previous slurpd replication).

I'm trying to test whether I can simply upgrade the master from 2.2.29 to 2.3.21
in place - possibly restarting the replicas. To do so I've set up the following
test systems (the two "sbuild" test systems on the right):

2.2.29 Master production ldap ----> 2.3.21 Slave | Master sbuild3 -----> 2.2.30 Slave sbuild5
| \
existing replicated slaves

where I'm trying to run the 2.3.21 server as a replicated slave to the current production
2.2.29 master and as a master to the test 2.2.30 slave. My hope is that we can test
replicating from the new master to the current production replicated slaves by just
pointing them to the new master with the current production content.

This 2.3.21 server seems to synchronize fine with the current 2.2.29 production master.
However, when I try to replicate to the 2.2.30 slave I don't get any data transferred and
I see the follow in the 2.3.21 "master"'s log:

May 24 13:21:42 sbuild3 slapd[8127]: slap_global_control: unrecognized control:

Might anybody be able to give me any hints about what that line is trying to tell me?

I don't see any transfer of the content for replication. With this same
configuration (just changing where the replication comes from) the 2.2.30 slave is able
to replicate from our current production 2.2.29 master server.

Are we going to have to upgrade our replicas before we can upgrade our master?
Is there anything I can do in configuration to get around this "unrecognized control"

--Jed http://www.nersc.gov/~jed/