IPTGuru.com

CUCM DRF Failures

Rob — Mon, 16 Jan 2012 23:59:09 +0000

Last week we had a RAID Controller fail one of our Cisco C200 UCS servers. It happens. For us it seemed to be just a mere annoyance at first, it took down two of our CUCM subscribers and our TFTP server. Phones failed over to anothe subscriber and our DR TFTP started serving up TFTP requests. After a quick a quick call to TAC Cisco had our replacement RAID controller to us in about an two hours. We powered the server down, replaced the controller and we were on our merry way. Or so we thought…

The next morning we started getting alerts from RTMT that our backs were failing for two of the CUCM servers that were affected by the outage. I tried the usual process of restarting DRF services on those two subscribers and kicked off a manual backup. Failure. Working with TAC for about 15 minutes they had it pegged. Corrupt File system on these servers. Solution? CUCM recovery CD. A quick fsck with the recovery cd and all was right with the world again.

Here is the error we were getting:

Reason : DRF was unable to backup component PLATFORM.Error : Unknown PLATFORM Error
AppID : Cisco DRF Master
ClusterID :
NodeID : ahpucm1020
TimeStamp : Fri Jan 13 09:13:53 CST 2012.
The alarm is generated on Fri Jan 13 09:13:53 CST 2012.

Hopes this helps someone who runs across this error in the future!

Basic Calling Order

Rob — Tue, 06 Sep 2011 01:36:34 +0000

This is the order in which calls are processed.

Phones/Gateways/etc –> Calling Search Spaces –> Partition –> DN

The most specific patterns are chosen regardless of the order or partitions.
Example: We have the following two partitions configured with the following two patterns:

Partition 1
5xxxx

Partition 2
50xx

We then dial 5001. Because Partition 2 has a more specific match it will be used to process the call.

[important]The order of partitions in not important unless there is a tie in the longest match for a DN. If there is a tie then the first listed partition containing the match wins.[/important]

CCIE Lab Study – Day 1

Rob — Wed, 27 Jul 2011 00:55:14 +0000

DHCP Pools for branch routers.

# Configure DHCP exclusions

ip dhcp excluded-address 192.168.1.1

# Configure DHCP pool

ip dhcp pool BRANCH1-PHONES

* Tip: Configure all options first before configuring network statements. This way you don’t have to reload phones.

# Configure TFTP (Option 150)

option 150 192.168.1.170 192.168.1.171

#configure default router

default-router 192.168.1.1

# Configure the network for the address pool

network 192.168.2.0 /24

International dial-peer Troubleshooting Tip

Rob — Sun, 24 Jul 2011 16:44:51 +0000

When using the “sh dialplan number” command to test/troubleshoot an international dial-peer you will need to add “timeout” at the end of the syntax. For example:

show dialplan number 9011123456789 timeout

Without the timeout command you will never hit the international dial-peer. This was news to me and the Cisco TAC engineer who had to look it up himself!

Cisco Live 2011

Rob — Sun, 10 Apr 2011 01:23:35 +0000

My company was extremely gracious enough to send me out to Cisco Live again in Las Vegas! What’s even better is that it will be the week of my birthday this year (July 11th)! If you are attending and interested in meeting up drop me a line!

rob@iptguru.com

UCCX 8 “Bug” to Watch Out For…

Rob — Sun, 20 Mar 2011 04:35:45 +0000

CSCtk35441 Bug Details basically states “Dont use admin account names that start with “UCCX”. From the 7.0.1 days our uccx admin account was….uccxadmin. I decided to use that in version 8 to keep things simple and it backfired on me. I couldn’t get in through CLI any more and I also couldn’t get into the OS Administration web portal any more.

Long story short, I am reinstalling!!! =)

Cheat Sheets Galore!

Rob — Mon, 03 Jan 2011 04:37:57 +0000

I found this link over at Ciskoblog.net and wanted to share! It features some REALLY useful cheat sheets for routing, voice and a few other networking goodies. Courtesy of packetlife.net

You can find them here

Importance of NTP

Rob — Mon, 06 Sep 2010 21:14:13 +0000

I had an issue last week regarding replication between my 8 CUCM nodes and couldnt get it work for the life of me. Low and behold after an hour of googling potential issues I found a post on Cisco’s support wiki saying that if NTP is not working correctly replication will fail. Well sure enough, I hop onto my publisher and execute a “utils ntp status” only get to a reply of “unsynchronized” because the source I had been using internally was having issues. I reconfigured NTP to point to an external source (thank you NIST) and replication was fixed and reporting status 2 within 5 minutes.

Lesson learned!

A Crappy BB picture of some new gear!

Rob — Wed, 28 Jul 2010 03:29:45 +0000

What you are looking at here is 3 Cisco MCS-7845-I3 servers. One for Unity Connection 8.0 and 2 for Cisco Unified Communications Manager. The CUCM cluster at this location will be made up of the following:

1x MCS-7835-H2 (Publisher)
2x MCS-7845-I3 (Subscribers)

The publisher will not see any phone registrations and the subscribers can handle a total of 15,000 phones total. But remember, this is only for this location! The CUCM will be geo-clustered across 3 different states over a dual carrier MPLS solution with more bandwidth than you can shake a stick at. I did beef up the publisher to 18GB of memory so that should help things along as well.

Sorry for the poor quality. Tomorrow I will break out the iPhone and get a better picture.

CCIE Written exam = PASSED!

Rob — Tue, 20 Jul 2010 18:24:43 +0000

I passed the voice written this morning and I am relieved that piece is over with now. Lots of work/study to do before the lab but I plan on taking that in the spring of 2011.