DBL Monitor - Max CDR Records not working

[m.gravel]

The publisher and all subscribers have the Max Records field set to 1500000 records and yet the CallDetailRecord table is currently holding 9670938 rows. The oldest entry is not being purged at night either.

The Maintenance Time is set to 24 and Maintenance Window is set to 2. This is the same settings we have for all our customers. All other customers have just a bit over 1500000 rows, and the old entries get purged nightly.

The only difference with this particular customer / CCM publisher is that we had to make the CDR database a publication at the customers request. The data from the CallDetailRecord table is pushed to an external SQL server on a regular basis.

Could this push have something to do with the ability of the DBL service to cycle out the old records nightly? I dont see any error messages in the event viewer application log at midnight when the maintenance window is scheduled nor at 1:00 am.

Any thoughts/suggestions would be appreciated.

[#2]

If we are using Max CDR Records as high disk space usage will increase a lot. You should really monitor the available disk space to make sure it does not get low. Things like remote StiBack operations during the night could also chew up unexpected amounts of disk space, so make sure you monitor over a period of time, not just during the day. If available disk space gets too low, you will definitely have problems (like DCD halting). Very large database tables can cause problems causing SQL transactions to hang. Specifically one case being if you use ART to perform automatic purges.

http://www.cisco.com/en/US/products/sw/voicesw/ps556/products_tech_note09186a0080100566.shtml

http://www.cisco.com/en/US/docs/voice_ip_comm/cucm/service/3_3_2/ccmsrva/sacar1.html