09-22-2010 07:31 AM - edited 03-14-2019 06:33 AM
We record every call into our UCCX 7 platform which makes my backups quite large. This causes an alert every night when the backups run from CRSAdmin stating that the server is running low on disk space and needs at least 40 GB free for B&R to run.
Is there a way to adjust this threshold?
Thanks
Solved! Go to Solution.
09-23-2010 01:39 PM
Hello Brett,
I see, thanks for clarifying that for me.
The UCCX installer set this threshold to a certain value based on the total disk space (not sure about the percentage thou).
There exist a registry key that holds the value of C: drive size
\HKEY_LOCAL_MACHINE\SOFTWARE\Cisco Systems, Inc.\CRS\DiskDrivesSize\C_DRIVE_SIZE.
You can manually change the value on this registry to make the UCCX believe you have a bigger HD and restart Node Manager. However I would only do that on a Lab environment, because this registry entry is also used to calculate some other things, for example the default size and number of some traces and logs files and other stuff.
You can also get rid of the message by just removing that email from the email subsystem
Pablo
09-23-2010 08:10 AM
Anyone? 5 stars for any info on this!
I don't understand why the alerting mechanism in UCCX is such a mystery. Come on Cisco...throw me a bone here!
09-23-2010 10:01 AM
Hi!
The UCCX Alarm Service uses xml-based catalogs:
C:\Program Files\Cisco\AlarmService\Catalog
You can enable the Alarm service by going into AppAdmin> System> Alarm Configuration.
This service will forward system events, warnings and errors to a Syslog server or to the standard Windows Event Log. You can find additional information in the Servicing and Troubleshooting Guide:
http://www.cisco.com/en/US/docs/voice_ip_comm/cust_contact/contact_center/crs/express_7_0/troubleshooting/guide/uccx70srvtrbl.pdf
Being said that, if you are getting an actual low on disk alarm, I think you probably have an alarm configured in Windows.
Please check by going into Performance Logs and Alerts in Windows Server 2003. You can add an alert based on a counter and whether it goes over or under certain limit. Like described here:
http://support.microsoft.com/kb/324796
Hope that helps!
Pablo
09-23-2010 12:12 PM
Pablo, thanks for the reply! I checked the event viewer on the UCCX servers and they are clean. This is not a Windows alert. This alert is coming from CRSAdministrator and is routed to the email address I have configured in the email subsystem (Appadmin / Subsystems / email.)
The server is not actually getting low on disk space when this alert is generated, it is simply going below the 40 GB free threshold that is built somewhere in the UCCX application. This is what the actual email looks like:
From CRSAdministrator@AUS-SVR-ICD-01
WARNING: CRS server AUS-SVR-ICD-01(172.21.X.X) is running low in free disk space
**DO-NOT-REPLY-TO-THIS-EMAIL**
CRS required that you have at least 43377659904 bytes available in drive C:, otherwise B&R will fail in operating. Please free up some disk space if required.
-Cisco Customer Response Application Administration.
So maybe the solution is remove the email address I have configured in the email subsystem for alerting and set up a syslog server instead for alarms?
Brett
09-23-2010 01:39 PM
Hello Brett,
I see, thanks for clarifying that for me.
The UCCX installer set this threshold to a certain value based on the total disk space (not sure about the percentage thou).
There exist a registry key that holds the value of C: drive size
\HKEY_LOCAL_MACHINE\SOFTWARE\Cisco Systems, Inc.\CRS\DiskDrivesSize\C_DRIVE_SIZE.
You can manually change the value on this registry to make the UCCX believe you have a bigger HD and restart Node Manager. However I would only do that on a Lab environment, because this registry entry is also used to calculate some other things, for example the default size and number of some traces and logs files and other stuff.
You can also get rid of the message by just removing that email from the email subsystem
Pablo
09-23-2010 02:18 PM
Ok I ended up putting in a dummy email address so that the email subsystem can still function as I have some alerts built into scripts for when calls come into a queue and there are no agents logged in to take them.
For better alerting I set up a syslog server and pointed the UCCX servers to it. This setup works much better. Thanks again for your help Pablo!!
Brett
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide