We continue to receive feedback on our recent series of newsletters about the lengthy Mean Time to Repair (MTTR) of application performance issues. We pointed out that in the vast majority of instances, the network is assumed to be the cause of any degraded application performance, even though it is the cause in only a small minority of instances.
We continue to receive feedback on our recent series of newsletters about the lengthy Mean Time to Repair (MTTR) of application performance issues. We pointed out that in the vast majority of instances, the network is assumed to be the cause of any degraded application performance, even though it is the cause in only a small minority of instances.
One reader wrote to us commenting on the lack of a coordinated approach to troubleshooting application performance issues. He stated that the process used to troubleshoot such issues ends up being a game of "pass the ticket," where a trouble ticket bounces from group to group with everyone adding in "my area looks ok." He says getting everyone in a room together for an hour produces great results. Often a server guy will mention something that causes the database guy to wonder about something else, which leads the application guy to get an idea that solves the problem.
He added that if you want to shorten the resolution time, increase the communication. He said that this approach is starting to be so effective at his company that they have designated a certain room as the "war room" that has enough space and network drops so that everyone can sit down with their laptops, use the whiteboard, and avoid their phone. He pointed out that it is not high tech by any means, but that it works.
We certainly applaud any attempt to increase the communications between and among the various components of the IT function. In addition, we are sure that this kind of communications will improve the MTTR application degradation issues and are pretty convinced that this solution is the best that the author could do given where he sits in the IT organization. However, this approach looks to us to be a Band-Aid to solving a problem that is being ignored by senior IT management. That being that application delivery is becoming increasingly important to the success of an organization at the same time that the IT organization is increasing stove-piped.
In this context stove-piped means that the individual components of IT do not have common goals, terminology, tools and processes. We believe that senior IT management needs to fix the IT organizational model at both a planning and an operational level. We will discuss the operational level of this new model in the next newsletter.
Further reading on this topic: Don’t assume application performance problems are always network-related; What makes application management so hard to do? ; NOCs now in charge of application management .