Cisco Support Community
Community Member

Unity connection 8.5 location requesting non existing dependent during replication


We have 2 cluster of Unity connection 8.5 both in HA (4 servers total). One cluster in city A et the other cluster in city B. Last week, replication between both clusters broke after the creation of a user on cluster A that had duplicated entries/conflicts in cluster B. We deleted the user from cluster A to bring the replication back, but the user never got deleted from cluster B. We went ahead and manually pushed the DB from A to B, problem solved, the user disappeared from Cluster B.

Here is where it gets interesting, since that event, Cluster B requests every 40 minutes or so that Cluster A provides information on a dependant object. A delay is present in the response from Cluster A, and a CUReplicator error shows up in RTMT with an PKID object number. That dependant object was traced back to be the EndUser PKID of the problematic user mentioned above. As you may guess, that user is nonexistent, we erased it. I queried the enduser table in the hope of finding a loose end, nothing on all clusters. I queried in the distribution list tables on Cluster B, thinking that the user may have been a member of a Cluster B distribution list,,,no luck.

So anybody has any idea how I can track a non existing dependence and stop Cluster B from asking for it?

05431af1-1cd9-4017-910c-6a082c888606 = PKID of our now deleted user, we made the corolation with the replication log bellow.

14:32:17.513 |9058,MessageProcessor,,CuReplicator,13,Received PUSH Replication: Location=mtlunc01 (push) USN(s)=4321 User=Etude10 Etudiant10, Objectid=05431af1-1cd9-4017-910c-6a082c888606 Time Stamp=1320461040925

Notice now how the user=null instead of Etude10 after we deleted the Etude10 user from cluster B with a push from Cluster A.

15:29:46.587 |9052,, dependent count=60),,CuReplicator,13,Dispatching: Replication: (pull) USN(s)=0 User=null, Objectid=05431af1-1cd9-4017-910c-6a082c888606

cluster B requests this every 40 minutes.

5:29:46.623 |9052,, dependent count=60),,CuReplicator,13,Sending changes for USN(s): 0, to mtlunc01 via message file CuReplicator_0_3249499062672782075.txt

15:29:46.623 |9052,, dependent count=60),,CuReplicator,13,Sent PULL for object 1:05431af1-1cd9-4017-910c-6a082c888606 to location mtlunc01

15:29:46.624 |9052,, dependent count=60),,-1,-1,Delay receiving dependent object05431af1-1cd9-4017-910c-6a082c888606 from location mtlunc01. Requests sent = 244. This situation may indicate network connectivity problems.

Could not track it down using those two queries on both clusters, the object is nowhere to be found.

run cuc dbquery unitydirdb SELECT Alias from vw_User WHERE ObjectId = '05431af1-1cd9-4017-910c-6a082c888606'

run cuc dbquery unitydirdb select * from vw_distributionlistmember WHERE ObjectId = '05431af1-1cd9-4017-910c-6a082c888606'


Everyone's tags (1)
CreatePlease to create content