Today I came across a CMS Database Replication issue that had me scratching my head for a while.
Scenario is 2 sites with SE and Edge each, the first site I setup a few months back and was expanding with the new site. For the sake of explanation I will call the existing site #1 and the new site #2.
All servers were replicating except for the #2 Edge Server.
I ran through the obvious checklist as follows:-
- Edge Server has a valid Certificate
- Edge Server can reach internal CA
- Edge Server is able to resolve Lync FE
- Can Telnet from FE to Edge on port 4443
(TIP: check the cert chains etc by browsing to https://edgeserverFQDN:4443/ReplicationWebService)
No joy..So I ran the XDS Trace on the Logging Tool. It was at this time that I realized that the usual 3 XDS trace components were missing as below:-
So where are the other 2 trace components?
It was at this point that the lights went on…the Lync Server I was tracing on (#2) was not holding the CMS! Embarrassing. So I launched the logging tool on FE #1 and what do you know:-
So what did the trace reveal? It showed how Edge Server #2 was not responding. So I suspected that although I had tested Telnet access from the #2 FE to #2 Edge – I hadn’t tested connectivity from FE #1 (where the CMS is) to #2 Edge.
My suspicions were confirmed, a quick chat with the networking boys and voila.