Cisco HP Call Manager Template

[Shmegles]

Here's a template for a Cisco Call Manager. We purchased our hardware direct from HP and I'm unsure if the HP specific items will work with Cisco OEM'd hardware. Graphs work on both DL380-G5 configured as 7835-H2 and DL380-G5 configured as 7845-H2.

Graphs show:
-number of gateways registered, rejected, and unregistered
-number of phones registered, rejected, and unregistered
-number of voicemail devices registered, rejected, and unregistered
-CPU temp (2 cores w/2 processors each)
-One min CPU usage for each processor
-Temp of some misc internals
-Uptime in hours

[churchi]

Awesome template. Just what i could use at my work.

Is this for the windows CCM 4.x and below? or for the new CUCM 5.x/6.x/7.x servers?

cheers

[Shmegles]

Good question...I made these templates for our 6.1 installation. I then verified that registered phones works on a 4.1 site. My guess is that they'll work with 6.x and 7.x, at least partially with 4.x and I'm unsure about 5.x. The HP graphs (CPU usage, temp, etc.) depend on the hardware you're using...IBM hardware (i.e. 7834-I2) won't display the HP stuff and I'm unsure about hardware purchased directly from Cisco.

Glad to hear the graphs are useful for you...I had lots of fun making 'em and they're working well for us.

[aleu]

Good question...I made these templates for our 6.1 installation.

Hmm, I am trying to use this template on my 6.1 installation. The template got imported just fine, I have configured the SNMP on my CM, but the graphs do not render (none of them). I have turned on the debug mode on one of the graphs and saw this:

RRDTool Says:
ERROR: the RRD does not contain an RRA matching the chosen CF

What does this mean?

Please advise.

[Shmegles]

A couple of thoughts for you...

1. Check out http://forums.cacti.net/viewtopic.php?t=3686 for a detailed discussion of this error. Try the fixes in there, such as changing the CF (a Consolidation Function) from LAST to AVERAGE. You'll have to do this for each Graph Template Item on each graph.

2. I'm happy to assist further, also. What version of cacti are you running? And can you post the entire graph debug output?

--Sean

[aleu]

A couple of thoughts for you...

1. Check out http://forums.cacti.net/viewtopic.php?t=3686 for a detailed discussion of this error. Try the fixes in there, such as changing the CF (a Consolidation Function) from LAST to AVERAGE. You'll have to do this for each Graph Template Item on each graph.

This did the trick! Thank you. Now I am facing another problem, as it seems that these templates are strictly for CM running on the HP hardware. My hardware is not HP and thus some of the graphs do not render 100% e.g. only two CPU cores rather than 4, as well as the last temperature sensor does not seem to exist on my platform.

Now really difficult question, how can I graph active calls? Is this even possible? It does not seem that CM exposes this information via SNMP...

Perhaps, this is a wrong direction and I should be looking for the templates that will graph my PRI router?

[Shmegles]

I, too, have yet to find a way to get active calls out of Call Manager itself. We are graphing active calls across each gateway, though and I just posted the template at http://forums.cacti.net/viewtopic.php?p=154819#154819.

If you have multiple gateways, I suggest installing the aggregate plugin and combining all gateway graphs into one stack graph. See above template link for an example.

[aleu]

I, too, have yet to find a way to get active calls out of Call Manager itself. We are graphing active calls across each gateway, though and I just posted the template at http://forums.cacti.net/viewtopic.php?p=154819#154819.

If you have multiple gateways, I suggest installing the aggregate plugin and combining all gateway graphs into one stack graph. See above template link for an example.

Thanks for the link, but it does not work for me. Well, I am able to pool the information via SNMP, but the data is incorrect. Take a look:
SNMPv2-SMI::enterprises.9.10.19.1.1.4.0 = Gauge32: 55684
and on another two routers:
SNMPv2-SMI::enterprises.9.10.19.1.1.4.0 = Gauge32: 0

Any idea?

[k-y]

I, too, have yet to find a way to get active calls out of Call Manager itself. We are graphing active calls across each gateway, though and I just posted the template at http://forums.cacti.net/viewtopic.php?p=154819#154819.

If you have multiple gateways, I suggest installing the aggregate plugin and combining all gateway graphs into one stack graph. See above template link for an example.

Sean, by gateways, do you mean calls coming into your IP Gateway application server?

[Shmegles]

I, too, have yet to find a way to get active calls out of Call Manager itself. We are graphing active calls across each gateway, though and I just posted the template at http://forums.cacti.net/viewtopic.php?p=154819#154819.

If you have multiple gateways, I suggest installing the aggregate plugin and combining all gateway graphs into one stack graph. See above template link for an example.

Thanks for the link, but it does not work for me. Well, I am able to pool the information via SNMP, but the data is incorrect. Take a look:
SNMPv2-SMI::enterprises.9.10.19.1.1.4.0 = Gauge32: 55684
and on another two routers:
SNMPv2-SMI::enterprises.9.10.19.1.1.4.0 = Gauge32: 0

Any idea?

Hmmm...that's definitely the correct OID to use but I'm unsure why you're getting such odd numbers. I suggest trying a reboot (I know, lame) in order to clear the snmp counters (which don't reset when a 'clear counters' command is issued).

Or, we can dig into the device more...what type of hardware and what IOS are you running?

[Shmegles]

I, too, have yet to find a way to get active calls out of Call Manager itself. We are graphing active calls across each gateway, though and I just posted the template at http://forums.cacti.net/viewtopic.php?p=154819#154819.

If you have multiple gateways, I suggest installing the aggregate plugin and combining all gateway graphs into one stack graph. See above template link for an example.

Sean, by gateways, do you mean calls coming into your IP Gateway application server?

By gateways I mean the actual hardware which does TDM to ethernet conversion. These devices handle conversion of calls from a VoIP environment out to the rest of the world (aka PSTN) or between VoIP and a PBX. In Cisco's architecture, these are small routers with PRI or POTS cards in 'em.

[aleu]

Or, we can dig into the device more...what type of hardware and what IOS are you running?

Sorry for the late response (I was visiting Japan ). The hardware I am using are 4 routers. All of them are Cisco 2811. First two run on ver 12.3 and the other two run version 12.4.

Thanks,
AL

[Shmegles]

Will you let me know specifically what version of IOS you're running? I'd like to know the entire string (i.e. c2800nm-ipvoicek9-mz.124-20.T1.bin).

[aleu]

Will you let me know specifically what version of IOS you're running? I'd like to know the entire string (i.e. c2800nm-ipvoicek9-mz.124-20.T1.bin).

Absolutely! Sorry for the much delayed response.
My routers run on c2800nm-advipservicesk9-mz.123-12.T5.bin.

[Shmegles]

Well, I don't know what why you're getting an odd value. I suggest a couple of things, in no particular order:

1. Open a TAC case if you have Cisco support
2. Reboot the router
3. Upgrade to 12.4 code
4. Upgrade to 12.4 'ipvoice' (if you can do without some of the advanced IP features)