You may have seen a write-up by Denise Dubie in Network World yesterday on our second annual “State of the Network Global Study.” We tried to understand the biggest challenges in troubleshooting and how much time was spent in troubleshooting.
The largest challenge in troubleshooting won’t surprise anyone who has been involved in networking; nearly three-fourths of respondents said “identifying the source of the problem” was their largest troubleshooting headache. As I looked around the web for other studies, that might be able to confirm our results, according to Zeus Kerravala, senior vice president of global enterprise research at the Yankee Group, in a Network World article, their “research shows that when troubleshooting problems, 90% of the time is spent identifying it.“
One of the reasons for performance problem causes being so difficult to determine might be a lack of information. According to our study, one-third of respondents indicated the need for better or more relevant information. Looking to other research in the area, Jean-Pierre Garbani, vice president and principal analyst at Forrester, said that the firm’s research indicates “that 74 percent of performance problems reported by end users are not detected by current infrastructure management.“
While troubleshooting challenges may be a no-brainer for anyone in networking, trying to understand how much time various troubleshooting steps can take is something I haven’t seen from other research. With that in mind, we broke down the conventional troubleshooting process into 5 very general steps:
- User reports problem
- IT waits for problem to recur or attempts to replicate it
- Start packet capture and determine the cause
- Analyze problem
- Resolve issue
Through the study, we attempted to identify how much time was spent in step 2 and step 3. We found that although a large portion of respondents (41%) spent less than 17 hours monthly (less than 25 work days annually) replicating issues, the majority spent above 17 hours monthly (25+ days) attempting to recreate the problem.
The other step of interest was #3, Determining the problem cause, we found most respondents spent 17 – 33 hours monthly (25-50 days annually). The graph provides a more thorough breakdown of the amount of time spent.
In upcoming posts, we’ll look at other trends in the State of the Network Study, and provide tips on network troubleshooting. For more info on the study, here are links to the study (PDF), press release, and a related article by Dave Bailey of IT Week.
April 2, 2008 at 9:21 am |
I don’t doubt that network admins are spending large quantities of time troubleshooting network issues. I do think that a large part of this is because many organizations have a lack of proactive monitoring of their networks.
It’s not that the tools don’t provide what’s needed, it’s that the tools aren’t implemented! Problems are handled when something breaks. This is a bad approach as the resources that you end up spending trying to fix things after they broke (not to mention possible lost revenue, e-mail, etc from the broken network) are a lot more expensive than the cost of getting good monitoring in place. Good monitoring helps stop problems while they’re still minor preventing the failure in the first place.
April 8, 2008 at 2:19 pm |
[...] to Decrease Network Troubleshooting Time In blogging about the amount of time our State of the Network Global Study found that IT staffs spent on very points of the [...]
April 15, 2008 at 1:40 pm |
[...] Confidence and Concerns I spoke earlier about the troubleshooting concerns and technology adoption rates detailed in our State of the Network 2008 study (view PDF). In addition to troubleshooting, we also [...]