UCCX 8.5 in HA Failover

[Stephen Keller]

So we tested our HA server today for about 4 hours and everything worked great.

A question i have though is what happens to the call statistical data that is generated while running on the secondary UCCX box?

We have an HTTP trigger that refreshes Queue stats like agents logged in , agents talking, average talk time, abandon calls, etc. to a web page.

This data was all refreshing normally, now that we have rolled back to the primary uccx box, all of those stats have been refreshed back to zero.

Will the call data that was generated while running on the HA server be replicated back up to the primary once it is online? I would expect that the data gathered during the HA time period would still be accessible to supervisors and HRC??

thanks in advance.

steve

[Gajanan Pande]

About

Stephen Keller wrote:
Will the call data that was generated while running on the HA server be replicated back up to the primary once it is online? I would expect that the data gathered during the HA time period would still be accessible to supervisors and HRC??

Data sync wont happen automatically, the admin have to do it manually, either using GUI ( Appadmin page ) or CLI. Manual data sync is indeed good because if it would have been automatic synch when both node comes online & imagine if the data size is huge, it would definitely load the call processing node & would bring down the contact center performance significantly . Having the manual option enables admin to do this sync activity during off-peak hours when there are no calls. Yes, once the DB is fully sync, HRC will be able to access all the data.

Refer the UCCX Admin guide to know how the DB sync is done in UCCX HA deployments.

Pls remember to rate helpful posts.

GP.

[Stephen Keller]

Thank you for the reply, I have been looking and cannot find where in appadmin pages this capability is. I have been in the admin guide for 8.5 and cannot find it there either. Do you happen to recall where this is?

thanks again!

[Ryan LaFountain]

Check the SRND for details about replication.

http://www.cisco.com/en/US/docs/voice_ip_comm/cust_contact/contact_center/crs/express_8_5/design/uccx851srnd.pdf

Your HTTP trigger is probably fetching out of the Engine's memory and therefore there is no failover of this information. The call and agent detail records are stored separately in the Informix DB for both CSD and HRC and this is replicated through Informix to the other node regardless of which node is active.

The active UCCX Engine that is processing calls writes the historicla data to its local DB and then it is up to replication to move it to the other side. This and what your HTTP trigger most likely uses are completely different processes when it comes to failover, data storage and writing.

[Gajanan Pande]

Am sorry, the option is on " Cisco Unified CCX Serviceability " page >> Tools >> Datastore Control Center >> Replication Servers

About CLI, " utils uccx dbreplication start " is the command.

Refer Page 47 of http://www.cisco.com/en/US/docs/voice_ip_comm/cust_contact/contact_center/crs/express_8_5/configuration/guide/uccx851cli.pdf

Pls rate the post if it helps.

GP.

　

　

　

[mysecure2017]

How did you manually failover your UCCX box? Did you just power them off?