Infrastructure Management Software Blog

Do you want to get rid of your event consoles?

A guest post by Mike Shaw.
- Peter


Are you in the "event consoles will live forever" camp or the "we want to get of event consoles as soon as possible" camp?


I ask because the world seems to be divided into two camps.


With events from top-down performance monitoring (user experience or tracking business transaction performance), it's relatively easy. You get a KPI violation and you turn the event into a ticket because you know it's service affecting -you are, after all, monitoring at the business service level. Once the incident has been created, however, you need to correlate against events to understand what is causing the business service to have problems (more on that from Michael Procopio in his posts on Problem Isolation).


But what about event(s) from below? Let's imagine you have a SAN, which is used by an active directory server, which is used by MS Exchange. Let's imagine the SAN has a problem . This generates an event for the SAN. The ADS complains and throws an event. The Exchange server throws an event too. We can do a number of event processing things automatically:


We can group related events and figure out what is the causal event. So, in our little example, we can infer that the SAN, ADS and Exchange events are an interrelated group and that the SAN event is the cause of all the trouble.


We can work out the business service impact -we can look upwards in the service dependency map and figure out that the Exchange server is used by N users. We can understand the SLAs resting on the Exchange server and how close we are to jeopardy on those SLAs. We can even figure out if any business processes are using that Exchange server. In other words we can automatically infer the business impact of our event stormlet.


So, we have enough to raise a ticket, don't we? We can raise the ticket against the top level affect business service(s) and we know that we are probably not raising tickets unnecessarily because we have grouped events and found the causal event. And we know the causal event and thus we know where the ticket should be routed.


Or am I being unrealistic? Are we not there yet -do you not trust the system to automatically raise tickets when you have a bottom-up event storm?


I would love to hear your views. Right now, my highly unscientific research has found that Europeans (home of ITIL ?) are more in favor of making the ticket the King/Queen and only using the event console to figure out what happened once the ticket has been raised.
One company wants to go even further. They want to get to the situation where they alwaysautomatically raise a ticket and then automatically route the ticket to the appropriate expert 2nd level group. Now, they may get the 2nd level allocation wrong some times, but if they do, they simply reroute to a 1st level triage group who manually reinvestigate who the ticket should go to. This is "manual triage by exception", if you like.


For HP Business Service Management, Mike Shaw.

Virtual Infrastructure Management (Q&A from HP-VMware webinar)

Thank you to everyone who attended the joint HP-VMware webinar on how to “Reduce Costs and Gain Control of Your Virtualized Infrastructure with Consolidated Management.” The speakers were Terry Lyons, Technical Alliance Manager, Enterprise Systems Management, VMware Corporation and Mike Shaw, Director of Product Marketing, HP.


If you missed the live event on October 7, 2009, you can view a replay here.


Here are the questions that people asked during the event, along with the answers.






































Question  Answer
Great to see you are working closely with VMware. What are you doing to support other hypervisors?  For now, the HP Virtualization SPI supports VMware ESX and Microsoft HyperV. We are looking at adding other hypervisors, based on our customer’s needs.
Can you explain the difference between siloed management compared to Central? What is the benefit to centralizing? Siloed management refers to managing each IT silo using a separate element manager.
Centralized management relies on a single event console to consolidate and correlate events from disparate IT domains. Using a single event console can reduce costs and speed the time to problem resolution.
How do you plan the systems and applications for the data center capacity?
Which kind of instruments/solutions do you use for the capacity planning? 
Please see our blog post on capacity planning, written by Hyperformix.

What is the minimum release of HP Operation Manager that enables the VMware API integration, or is the VMware SPI the only dependency?  First, the Smart Plug-In (SPI) is called the Virtualization SPI, because it supports hypervisors beyond just VMware, although that is the most commonly requested platform.
The Virtualization SPI works with Operations Manager on Windows 8.10, Operations Manager on Windows 8.16, Operations Manager on Unix 9.0, and Operations Manager on Linux 9.0.
Can you talk about how HP SIM fits in to this? i.e. hardware and VM Management. thanks,  HP Operations Manager consolidates events from all the various element managers across the enterprise. In the scenario you mention, SIM will send its events to Operations Manager where a single team can monitor the HP servers and any other infrastructure. There is a SIM Smart Plug-In (the SIM SPI is available at no charge) that integrates between SIM and Operations Manager. If an operator needs to perform advanced troubleshooting of the HP servers, he or she can launch SIM through the Operations Manager console.
Can this solution allow me to generate FTE savings?  We answered this during the webinar. This solution centralizes event management to a single Operations Manager console, leaving vCenter as the expert tool for advanced troubleshooting or escalations. This shifts event management to your tier 1 operators in the Operations Bridge, freeing your virtualization administrators (very hard to find these days) to work on more strategic tasks.
What specific HP software products are needed to implement the scenario presented?  We answered this during the webinar. You would use Operations Manager, agents for each of the managed nodes, along with the Virtualization SPI to manage the hypervisor.


 For HP Operations Center, Peter Spielvogel.


Get the latest updates on our Twitter feed @HPITOps http://twitter.com/HPITOps


Join the HP OpenView & Operations Management group onLinkedIn.

Search
Showing results for 
Search instead for 
Do you mean 
Follow Us


HP Blog

HP Software Solutions Blog

Labels
The opinions expressed above are the personal opinions of the authors, not of HP. By using this site, you accept the Terms of Use and Rules of Participation