This document answers frequently asked question about Jabber Service Discovery using User Data Service (UDS).
FAQ from Live Webcast
1.User Data Service (UDS) is currently based on Unified Communication Manager (UCM) database information, firstname, lastname and phone number. Is there any plan to increase the available fields and search options (for example, by department.)
A: We will have to check the roadmap with the business unit (BU). Currently, as you rightly mentioned, it does not have support for department field.
2.What about Jabber behavior when client has chosen email as the UCM userID ? It has 2 @ into the Jabber userID. For example my userID is email@example.com and my presence domain is also set as top domain : techies.com, Assume, my IM address is firstname.lastname@example.org@techies.com.
A: Jabber uses the <domain name> field after the @ to send the service record (SRV) query. So if the UCM userid is a top-level domain (TLD) email address (which is same as the TLD, it just uses the same name to get the SRV record from the dns.
3.Is it mandatory to enable the "Home Cluster" parameter in the end user page to get it work ?
A: Yes it is required to set the parameter in the end user page. That is the base configuration, and on basis of that the UDS will query and get the location of home cluster in real time.
4.I think first SRV lookup for @example.com points to WebEx Connect cloud service. Right? What is the right migration path to migrate from cloud service to On-Prem deployment?
A: Yes the first SRV sent by Jabber is always to check if the registered domain is a WebEx domain or not. If it is indeed a WebEx maintained domain, then Jabber does not query for other SRV (_cisco-uds and _cuplogin).So to migrate from Cloud to On-Prem, there is a command line interface (CLI) installation mechanism introduced in Version 9.7(1) to exclude Cloud services. This installation type forces the client to use _cisco-uds SRV and it will not check to see if the domain is a WebExregistered one and the installation to accomplish this is: msiexec.exe /i CiscoJabberSetup.msi CLEAR=1 EXCLUDED_SERVICES=WEBEX . But for pre Version9.7(x), I believe the only way as of now is removing the domain from WebEx administration site.
5.When will Jabber 9.6 be doing the service discovery which will be primary query cuplogin or cucm_uds ?
A: The order is this: 1. Webex, 2. _cisco-uds, 3. _cuploginso _cisco-uds will be sent ahead of _cuplogin. _cuplogin is not required if the UCM version is 9.x and above. But client will send _cuplogin also.
6.What is the SRV record _jabber._tcp on port 0 intended for ?
A: Basically it is used to enable a feature called Extensible Messaging and Presence Protocol (XMPP) Federation between Cisco Unified Presence and other clients like Google gtalk and IBM Sametime. For example, a customer that is using presence server and Cisco Unified Personal Communicator/jabber (CUPC or Jabber) and they want to enable federation (to see the presence status and do chatting) with another client like google talk (gtalk) or IBM Sametime client, then on the DNS server that resides on CUPS side, a SRV has to be created like this: _jabber._tcp.gmail.com you see the name gmail.com?
Simply, this is a SRV required on the public DNS to enable XMPP federation between Cisco Unified Presence, gtalk, and IBM Sametime.
7.Support of Computer Telephony Integration (CTI) and LDAP Integration via collaboration edge?
A: Currently Deskphone control (CTI) and Directory search against LDAP are not available in Collaboration Edge. Also, only UDS search is available in collaboration edge mode.
8.When we have multiple Cisco Unified Communications Manager (CUCM) clusters within a company, we set on each the same top domain. And do we need to set on each cluster a fully qualified domain name (FQDN)? Is there a relationship between UDS/Intercluster Lookup Service (ILS) and cluster FQDN parameter?
A: Yes, the FQDN is required for this topology you discussed.
9.Supply an overview on Global Dial Plan replication.
A: Global Dial Plan Replication (GDPR) uses the ILS to share dial plan information between participating ILS-enbaled cluster, GDPR allows each cluster to distribute information about its associated URIs, +E.164 numbers, enterprise numbers, +E.164 patterns, enterprise patterns, and Public Switched Telephone Network (PSTN) failover numbers. Each participating cluster shares a common Global Dial Plan. This link can be used for reference:
10.I tried to setup ILS using Transport Layer Security (TLS) authentication but it ís taking too long in getting listed in the topology.
A: Check for any alarms in the Real-Time Monitoring Tool (RTMT). TLS authentication requires certificate consolidation of participating nodes. Ensure that all the clusters have exchanged tomcat certificate prior to starting the ILS Configuration.
11.Is there any ILS Architecture change moving from Version 9.x and 10.x?
A: Yes. In Version 10, ILS only runs on Pub. The reason is with the global dial-plan replication (GDPR), ILS has to write to the database and that can only be done on the Publisher. ILS is redundant in that all hubs are fully meshed.
12.Can you explain the dependency between ILS and UDS?
A: UDS relies on the list of detected remote ILS clusters displayed in Advanced Cluster overview. UDS periodically checks this table for all detected remote ILS clusters and then accesses UDS resources on these clusters to find out about the UDS nodes in the remote clusters. Hence service discovery transactions do not rely on ILS. These are UDS transactions that use persistent state. Identify the nodes, which are running, and UDS and validate by running user query using Web browser.
13.Is data learned from ILS lost when the publisher is down?
A: No. The ILS data is persistent and remains.
14.How does UDS redundancy work?
A: Depends on DNS SRV implementation.
15.Why can I not I find utils find xnode in CUCM 10.x?
A: CUCM 10.x and later ILS runs only on Publisher and there ís no xnode concept anymore. Without a Publisher server, ILS will fail to establish connection with Hub/Spoke node.
16.Will UDS work if ILS is non-functional?
A: Yes. It will continue to work as long as the advanced configuration page has the configuration of the Nodes with UDS activated.