Hi,
Most of the times, the error is caused because of a name resolution problem on the client machine or a network reachability issue. But, I suspect a DNS name resolution as a potential reason for your issue.
A few troubleshooting steps that you can try:
1. If the _cisco-uds SRV is configured as either hostname or FQDN of CUCM instead of IP address, check if the server is resolvable against the DNS. This is called A record lookup. Just type nslookup <FQDN of CUCM> and nslookup <FQDN of IM&P> as configured in the SRV. Remember that you can't resolve plain hostnames on a windows machines without appending the domain name. There must be an A record for the CUCM and IM&P servers as FQDNs in the Forward Lookup Zone of the DNS.
2. If you get good results from step-1, i.e. the servers are resolvable agains the DNS, then check the status of services on the IM&P via CLI. You can run 'utils service list' and see if there any services in 'Not Running' state.
Check if you have a domain name and DNS IP configured on CUCM and IM&P via CLI using 'show network eth0 detail' command. If they are, then the servers can be resolved with their FQDNs against the DNS as long as the A records exist. Even if the servers do not have a domain and DNS server configured, Jabber client sends a DNS name resolution query for the hostnames of CUCM servers which fails. The reason for this is that the client gets the list of CUCM servers with their hostnames during the login process. And, it should be able to resolve that name failing which you will see the error.
Hope this helps!
-Sankar
