Do you have an ITIL problem manager?

Michael_Procopio — Sun, 04 Oct 2009 04:38:00 GMT

by Michael Procopio

What do Healthcare, Banking and Managed Service Providers have in common? Well at least the ones I spoke to all had ITIL problem managers. I’ll define problem management then give some examples.

What is the role of an ITIL problem manager? This was role introduced in ITIL v3 as part of Continuous Service Improvement. To understand the role we need to back up and describe incident and problem management.

An incident is a service interruption. Incident management is the process of restoring normal service. Frequently this is done with a work around, like killing a process and restarting or rebooting a system. Incident management is not worked about finding the root cause of the interruption.

The goal of problem management is prevent incidents from happening. What this means in practice is finding the root cause of incidents and fixing them and maybe along the way fix some others before they happen.

Let’s move to a couple examples.

1. Those darn log files. This came out in the first interview I did. What this showed me is that problem managers work closely with folks working on incidents. The result of the incident investigation was that an application written in-house wasn’t checking for disk space before writing the log file. Result, volume ran out of space and application stopped working. This was an easier find than many, but the problem manager put this in her monthly newsletter to developers and QA teams to prevent future problems.

2. One thing software can not fix is hardware. You have mission critical application and you end up getting hardware problems approximately every month for a few months. The problem manager, as part of his normal duties, is constantly monitoring incidents looking for patterns. While looking through incidents notices this pattern and starts investigating. In this case after talking to a number of folks and not hitting any answer he started investigating the hardware records and found the hardware was over four years old which is over their policy standard. Somehow the systems supporting this application missed their refresh. As you might guess once the hardware was refreshed the problems stopped

3. Sometimes its what’s not happening. A global ecommerce capability is not working at one location for 8 minutes. This problem manager says he is in the subfield of epidemiology, I think of more as forensics. This brings up a second learning I had: problem managers often orchestrate the process as much as they investigate themselves. After looking at a number of things that were happening to no avail, he asks each team to look at log files for abnormally low activity. This took some convincing, but the network team found a router that was showing almost no activity relative to normal. I didn’t get all the details but the source was someone made a change to a routing table and 8 minutes later changed it back. That change routed traffic around the site in question. The problem manager made an interesting observation: network folks often think what they do won’t affect anything and since they rarely get feedback to the contrary it re-enforces that opinion.

I’ll wrap of this post with what we can do to help the problem manager. For workflow we have the HP Service Manager Problem Management module. This includes logging, categorization, prioritization, communication and progress tracking.

For investigation we have HP Problem Isolation. Its proactive module helps find anomalies in performance data that can indicate potential upcoming problems. It also brings together incident and change data to help determine the cause of the problem.

Advanced analytics reduces downtime costs – detection

Advanced analytics reduces downtime costs – isolation

HP Software Universe - Mainstage Robin Purohit

Michael_Procopio — Thu, 18 Jun 2009 01:19:00 GMT

After a break Mainstage split into a Business Intelligence session and a Business Technology Optimization (BTO) session, I went to the later hosted by Robin Purohit, vice president and general manger, Software Products, HP Software & Solutions

Here are some bits I found interesting:

There are 15M virtual servers now and it is expected to double in 2 years.
Three key areas of change IT will see are: virtualization, SaaS and Web 2.0.
Web 2.0 screws up a lot in management, HP is building Web 2.0 compatibility into BTO tools.
Dataprotector, monitoring tools and mapping tools are including or working to include virtualization capability.
Service Manager and Asset Manager are now available as SaaS.
35% of flash apps violate Adobe best practices. HP released a testing tool which you can download for free called SWFScan. The same technology is built into some of HP’s BTO products. Customers can also do a free trial of our scanning SaaS.
One customer cut $5M in costs just in automating incidents with HP Operations Orchestration.
HP Software recently announced a suite of software for operations of Blackberry implementations. See HP and RIM Announce Strategic Alliance to Mobilize Business on BlackBerry for more details.
HP Software is publishing configuration management best practices later this month.
Brian Byun VP Global Alliances from VMware came on stage and announced announced an expanded partnership with HP to jointly develop software to manage VMware hypervisor technology. Denise Dubie at Network World wrote HP, VMware team to manage virtual servers.

There are a variety of Twitter accounts you can follow as well as the hashtag #HPSU09. Search Twitter for #HPSU09.

HPITOps – Covers BSM, ITFM, ITSM, Operations and Network Management

HPSU09 – show logistics and other information

HPSoftwareCTO

informationCTO

HPSoftware

BTOCMO – HP BTO Chief Marketing Officer

For HP BSM, Michael Procopio

Application Management : Service Manager

Do you have an ITIL problem manager?

HP Software Universe - Mainstage Robin Purohit