Cisco Jabber 10.5 users show offline

[saship]

Folks, 

I have an interesting issue with Cisco Jabber. 

Here is the setup - 

CUCM - 10.5 Cluster 

1 x Pub

2 x Sub's (Voice & Video) 

2 x Sub's (IM & Presence) 

Jabber for MAC version - 10.5.1 

When all the users are assigned to the same Presence Server the users can see their Presence status (Online, Offline etc), but when users are assigned to difference presence servers within the cluster they cannot see their status, the users show as offline even though they are online. 

Anyone seen this issue before ? I am hoping this is just a configuration issue. 

[Nirmal Issac]

Is DBReplication fine between the servers? Do you see any errors in the cluster topology page? If not, could you try restarting XCP Router service in both the nodes?

[saship]

Nirmal, 

The Db replication does look fine, I have included the output from both servers. btw here is the Layout of the cluster, the CUCM cluster is comprised of 5 servers, a CUCM Pub, 2 x CUCM (voice & video) Sub's and 2 x CUCM (IM & Presence) Sub's as shown below.  

CUCM (Voice & Video) - Pub / TFTP 

CUCM (Voice & Video) - Sub 1 

CUCM (Voice & Video) - Sub 2

CUCM (IM & Presence) - Sub 3 (also the presence Pub) 

CUCM (IM & Presence) - Sub 4 

Db replication status on CUCM Sub 3 (which is the presence Pub) 

admin:utils dbreplication runtime state

Server Time: Sun Jan 25 10:30:17 CST 2015

Cluster Replication State: Replication status command started at: 2015-01-25-10-29

     Replication status command COMPLETED 200 tables checked out of 300

     Last Completed Table: alarmconfig

     No Errors or Mismatches found.

     Use 'file view activelog cm/trace/dbl/sdi/ReplicationStatus.2015_01_25_10_29_38.out' to see the details

DB Version: ccm10_5_1_10000_9

Repltimeout set to: 300s

PROCESS option set to: 1

Cluster Detailed View from TX1UCMSUB03 (2 Servers):

                                      PING      DB/RPC/   REPL.    Replication    REPLICATION SETUP

SERVER-NAME         IP ADDRESS        (msec)    DbMon?    QUEUE    Group ID       (RTMT) & Details

-----------         ----------        ------    -------   -----    -----------    ------------------

TX1UCMSUB03         10.1.32.13        0.019     Y/Y/Y     0        (g_6)          (2) Setup Completed

CA1UCMSUB04         10.2.32.13        0.242     Y/Y/Y     0        (g_7)          (2) Setup Completed

Db replication status on CUCM Sub 4 (which is the presence Sub)

admin:utils dbreplication runtimestate

Server Time: Sun Jan 25 08:30:20 PST 2015

Cluster Replication State: Only available on the PUB
DB Version: ccm10_5_1_10000_9
Repltimeout set to: 300s
PROCESS option set to: 1

Cluster Detailed View from CA1UCMSUB04 (2 Servers):

                                      PING      DB/RPC/   REPL.    Replication    REPLICATION SETUP
SERVER-NAME         IP ADDRESS        (msec)    DbMon?    QUEUE    Group ID       (RTMT) & DB Status
-----------         ----------        ------    -------   -----    -----------    ------------------
TX1UCMSUB03         10.1.32.13        0.299     Y/Y/Y     0        (g_6)          (2) Setup Completed
CA1UCMSUB04         10.2.32.13        0.023     Y/Y/Y     0        (g_7)          (2) Setup Completed

 Also here is the topology view 

This is the Presence Publisher 

Here is the presence Subscriber 

Also you will notice that I have all 8 users assigned to TX1UCMSUB03 because if I move users to balance them across both servers, I lose the ability to view presence status 

Here is an example 

when all users are assigned to TXU1UCMSUB03 as shown below

I can see agent 1's presence status as available

When I switch Agent 1 over to CA1UCMSUB04 server as shown below; 

Agent 1 shows as offline 

even though Agent 1 is logged in and Available. You will also notice that Sashi (my account) shows offline to agent1, so basically if users are assigned to separate servers in a cluster they show offline to each other. 

I have tried rebooting the servers, restarting XCP router service and besides, I do not have any errors or warning on my diagnostics page or troubleshooter page. Everything looks healthy, I wonder if there is a setting that I missed or is misconfigured. 

[Marcelo Morais]

Hi Saship,

 just to double check ...

1) in CUCM > System > Service Parameters > Cisco CallManager, you configure the "Default Inter-Presence Group Subscription" as "Allow Subscription", am I correct?

2) in CUCM, in the 'SIP Trunk' configuration to IM&P, could you please send the info of the 'Inbound Calls', 'Outbound Calls' and 'SIP Information' window?

[jpcano_col]

Hello Saship and mmorais, 

I have the same problem. I would like to know the resolution. 

On my system the "Default Inter-Presence Group Subscription" as "Allow Subscription" This is correct.

On the second matter here are the screen shot.

[Marcelo Morais]

Hi jpcano_col,

 in the 'Inbound Calls' window > CSS > enter 'CSS_BHISC' and try again.

[saship]

mmorais,

I checked the service parameters and also attached below is the screen shot of the information you requested, as far as I can see everything looks normal

Here is the SIP trunk between CUP & UCM,  I have 2 trunks since I have 2 CUP servers and both SIP trunks are configured exactly the same with the exception of the IP address

The GTWY-CSS pretty much has access to every partition on UCM.

Here is the SIP information

Here is the SIP Security profile if interested

and the SIP Profile as well

[Marcelo Morais]

Hi saship,

 you must only configure one 'SIP Trunk' between a CUCM Cluster and an IM&P Cluster.

 After you configure the 'SIP Trunk', you must assign it as the 'IM and Presence PUBLISH Trunk' on CUCM Administration > System > Service Parameters > Call Manager.

 Make this change and try again.

Hope this helps.

[saship]

mmorais,

Thank you for pointing out that you only need 1 SIP trunk. I went ahead and reviewed the configuration guide and removed the second SIP trunk and modified the first one to use DNS SRV and set up the SRV records on my DNS server as shown in the configuration guide. At this point there is a single trunk pointing to a DNS SRV address which in turn is configured on my DNS server to resolve to the 2 Presence nodes for the SIP service

and the SIP trunk is assigned in CUCM service parameters

still no go, the problem persists.

[Nirmal Issac · Accepted solution]

Thank you for the detailed explanation (+5)

 I would like you to do one more test.

- Could you assign 2 users in the the second node and see if the presence works between them?

Basically the issue is with the propagation of Presence status between the servers. Could you confirm if the 'Router-to-Router' mode is selected as the communication mode between them? MDNS creates problem is the network is not configured for multicast.

After checking the above setting, please open a TAC case.

[saship]

Nirmal,

I logged in as 2 users on node 2, they can see each others presence status just fine. So it is just Inter node presence that is messed up. With regards to the Router-to-Router Mode I assume you are talking about the Security settings "enable XMPP Router to Router Secure" mode. This is enables as shown below. If this is not the setting please let me know if there is a difference setting that I need to enable.

[saship]

Nirmal,

after I posted my response, I started looking around and found what you were describing about Router-to-Router Mode vs Multicast DNS (MDNS) mode. once I changed the routing communication type under Presence administration > System > Service Parameters > Cisco XCP Router > Routing Communication Type as shown below

Previously set to Multicast DNS

Now set to Router-to-Router (Auto)

Everything works perfectly now. Thanks again for your help, the issue is now resolved.

[saship]

Nirmal,

after I posted my response, I started looking around and found what you were describing about Router-to-Router Mode vs Multicast DNS (MDNS) mode. once I changed the routing communication type under Presence administration > System > Service Parameters > Cisco XCP Router > Routing Communication Type as shown below

Previously set to Multicast DNS

Now set to Router-to-Router (Auto)

Everything works perfectly now. Thanks again for your help, the issue is now resolved.