In my younger days (many decades ago), I learned how to troubleshoot through a program to fix a particular issue. WAY back then, in the days of punch-cards, troubleshooting required logical and systematic thought. We couldn’t simply tweak the code, recompile, and try again five minutes later, because the process to submit changes to keypunch, receive the updated program cards, and get time on the compiler could be hours.
I think I was a more patient problem-solver back then.
Today, we expect instant solutions. It’s unusual to have the same problem vex us for over a week. We tend to either fix quickly or find a work-around and move on.
This past week we had a strange problem that appeared to be related to Exchange autodiscover service. We had just upgraded our Exchange server and a handful of people were having trouble with automatic replies and shared calendars. We tried several fixes connected to autodiscover, but to no avail.
After a week of intermittent work on the problem (with no real results), I decided to tackle it like I would have years ago…logically and systematically.
First, I documented the symptoms — ALL of the symptoms. Turns out the problem was affecting not only autodiscover services in Outlook, but it was also causing sporadic webmail connection issues in Chrome and Firefox (IE was not affected).
Next, I documented what I had tried so far. It included autodiscover settings as well as more network-specific stuff like flushing DNS, renewing DHCP lease, disabling IPv6, etc.
Here’s what I find amazing… Within 15 minutes of documenting the symptoms and trials/errors, I had solved the problem! Why?
The fundamental goal of problem solving is to find a solution. Within the context of IT (and many other disciplines), problem solving often includes eliminating possibilities. We try this and that, hoping that this or that will fix the problem. When it doesn’t, we try something else.
But without logic and organization, we can quickly lose track of the relationship between the various symptoms and trial/error attempts. We may shift focus away from the primary problem. Or, we may forget to try things that are known to have been problematic in the past.
So it was with our “autodiscover” problem. Years ago, disabling anti-virus was a go-to solution for MANY strange network connection problems. I haven’t hit this issue in years. Yet, until I made a logical list of my trials/errors, it didn’t occur to me to disable AV to see if the problem persists. In our case, it was a known issue with a very specific version of Kaspersky client. I should have thought of this earlier in the process (I know, you’re thinking “what a newb!”). But until I organized all of the facts, it really didn’t occur to me.
Lesson learned: next time you are vexed with a problem – of any type – take the logical approach and make a list of possible solutions, record what you’ve tried in the past, and use this information to narrow down your options. This approach helps with tech problems, as well as business opportunity decisions, staffing problems, budgeting issues…the possibilities are endless!