Jabber - IM & Presence - cisco synch agent service stopping automatically

[mohsin majeed]

Hi,

I am facing many issues in presence server after deployment. All error are attached.

- dbreplication problem

- Cisco sync agent service stops automatically on IMP node1 within 10 minutes after i manually start

- Many services are stuck in "Starting" on IMP node 2

- Failover doesn't work at all

- When i check all Services to activate in Serviceability, after some loading services don't activated

- Only good thing that users assigned to node 1 and Jabber can register and call incoming and outgoing. But keep loading the msg "Setting Phone"

- Tried many reboots but in vain.

- Now i shutdown the 2nd node and looking for solution :(

Deployed ISO/OVA from data store which shipped with BE6K

IMP

- cucm_im_p_11.5_vmv8_v1.2.ova

- UCSInstall_CUP_11.5.1.12900-25.sgn.iso

CUCM

- cucm_11.5_vmv8_v1.0.ova

- Bootable_UCSInstall_UCOS_11.5.1.10000-6.sgn.iso

Note: Should i think about to deploy MRA when IMP is not in good state; but only the jabber is working for incoming and outgoing call.

Errors and Information is attached

Regards,

[Slavik Bialik]

First of all, why didn't you enable the services on Node #2? You must do it, especially you're trying t run a cluster of IM&P servers.

Second, can you elaborate on your networking architecture of the UC servers? Because I see that both servers (and I'm guessing the CUCMs) aren't in the same network, means that I'm guessing there's a Firewall in the middle. And if that's the case, you probably experiencing issues because some of the relevant ports are blocked between IM&P - Publisher and IM&P - Subscriber. You must check it out, if you really have a firewall between those 2 LANs: 10.10.111.0/24 and 10.10.110.0/24.

Follow this port usage guide for IM&P in order to verify that all the relevant ports are opened:

http://www.cisco.com/c/en/us/td/docs/voice_ip_comm/cucm/admin/11_5_1/sysConfig/CUCM_BK_SE5DAF88_00_cucm-system-configuration-guide-1151/CUCM_BK_SE5DAF88_00_cucm-system-configuration-guide-1151_chapter_01010101.pdf

If you're experiencing so much problems, it must be the network and closed ports I think.

[mohsin majeed]

Hi Slavik,

Thanks for your time,

- I tried to activate these services on Node 2, but they don't activate, some stuck in "Starting" and some simply don't become active.

- On Node 1 all services starts except cisco sync agent stops after within 10 minutes after activating it each time.

- Yes both UCS server are in two different buildings with different subnets but very near locations with fiber connectivity in between. Each UCS is configured with Pub and Sub combination

- Network is very stable but about the firewall hopefully you are very right as i never thought about it.

============================

- Before reading your comment what i did it mess today; (searching from the forum i thought, may i have deployed wrong template which is 500 Users Full UC and it is not listed in 11.5 deployment as per

http://www.cisco.com/c/dam/en/us/td/docs/voice_ip_comm/uc_system/virtualization/virtualization-cisco-ucm-im-presence.html

i decided to build IMP nodes from scratch with 1000 users full UC which is listed in BE6000M - S TRC deployment. So, I deleted Node 2 virtual machine and from the cluster also. System>Server. But, when i tried to delete Node 1, it ask to remove from presence redundancy group; when i try to delete this group, it don't allow me to delete with message that " IMP can't be deleted while users are assigned to it". But, i already unassigned all the users from IMP server. I checked the dependency records, there are 0 items associated with server and presence redundancy group, still not deleting the Node 1. See Attached

This is critical client giving not much time, if i read your post in the morning then sure i would go to check firewall with client.

Now the situation is Node 1 is still there but node 2 gone coz i deleted it from the disk. As i now node 2 made me crazy when i was installing it. It took me 30 minutes to type this story, hopeful you will get me.

Now, the suggestion and help required to proceed further.

Best Regards,

[Slavik Bialik]

Heh, I totally get you.

And it makes very sense that this is a firewall issue, because it looks like the services cannot communicate with each other, which can explain for example why the services on Node #2 cannot be started.

Just to prove a point to your client, ask him to open any-to-any in the Firewall between all of the UC servers (CUCMs and IM&Ps) in bi-direction traffic, and it'll probably start working.

Me for example, if it's a big project, I always start when all the ports are opened, I mean, any-to-any opened. Only when I complete everything, and everything is working smoothly, I apply to Firewall policy and I'm doing few tests again to see if nothing is damaged because of the firewall. So I suggest you do the same.

BTW, installing Expressways with MRA, most of the job here is firewall, heh. You have the internal network, and the DMZ area where the Expressway-E resides. You'll probably will go for the single NIC topology, so you'll have to ask your client to implement a NAT reflect on his firewall and etc.

Anyway, good luck, hope it'll be all sorted out ;)

[mohsin majeed]

Thanks slavik for your suggestion, will work out and get back with update.

[mohsin majeed]

Hi Slavik,

I checked with client, there is no firewal between the subnet of IMP Node1 and Node2. Now again, my mind is going back as i deployed template with 500 users full UC which is not listed in the document for 11.5 deployment. I would like to deploy template with 1000 user full UC listed in the document.

Now the issue is, i am not able to delete the Node1 server from Presence redundancy group and System>Server. The error while deleting that IMP can't be deleted when user are assigned.

I double check, no user are assigned in Server, Presence Redundancy group, User management and IMP server.

How i can delete IMP servers in order to prepare them from scratch.