top of page
  • Writer's pictureLandon Foster

Where do I start?

I'll say it once, and I'll say it a million more times: Troubleshoot UP the stack, and you'll get it every time. With my Jr Engineers and support techs, past and present, I've always given the same advice; It's a better workflow.

Firm Handshake, Eye contact, Always troubleshoot up the stack

Everyone can troubleshoot a layer one issue (That is, physical). Well, Everyone except your customer. Over years of troubleshooting in many capacities, what I've found most often is what just about any network engineer will tell you and has become almost boorish to mention: Try turning it on and off again. If that isn't possible, or just didn't work, lets get a bit more complex- Is it plugged in? Now, the more senior above you might think this is blindingly obvious, but you have to start somewhere and a lot of issues are fixed here. Honestly, it can feel like magic, but the thing is, people smarter in their respective fields have designed a lot of the gizmos we use every day and built them to work and check themselves from a reboot; Even in extremis, a hard reboot. This can also include simply asking someone to take a look at indicator lights- They're there for a reason. A lot of us get excited about an esoteric issue and rev up the Pcap... but that doesn't help if there's nothing to capture. That brings us to the next most common - Most of you reading this will be more of the networking persuasion, and the next one comes for you; Ready your field tech cursing.

From IT Crowd- Not mine.

Is it plugged in? The amount of time I've spent troubleshooting mediums, even in wireless, and found that the ethernet or other bounded medium wasn't seated correctly is kind of sad. Allegedly, one of those times was spent cursing vehemently in the rain in the middle of a Lousiana swamp. This is one of the irritations that all the technology in the world won't fix. Cables wiggle- why, and how is something we may never know. It doesn't matter if you're looking at old CAT3, new Cat6, or some fancy Fc/Sc type fiber. They all do it, and your time is best served giving it a go. Obviously if you're getting power and traffic steadily, you can skip this- Move on to layer two. But now we've established that it's on and it's in tight, and no gremlins have managed to swap out all your cable ends with millions of ants, and that's progress. (Yes, this absolutely happens. "CrazyAnts" are absolutely a thing, and in the US south the really, really like getting into any base station modules they can get their dirty little antennae on.) Last thing to check as far as most common in complete network failures is with ethernet. Are you using the right cable pin-out? Most NICs these days will compensate just fine, because of magic. I refuse to believe it's anything else, but some NICs aren't blessed with this wizardy, so take a look anyway, yea?

Cat 5 Pin-out, Tab-down (above) Note 568B is almost universal

To close out the more basic layer 1 stuff, lets look at a sneaky one. Your AP might work, or it might go on and off. Check your power budget if you're using PoE! Most APs these days vary the power it drags in somewhat dynamically, so you might be exceeding your budget. A power budget, put simply, it how much your PoE enabled switch (I hope you enabled PoE...) will put out. Be careful that you've taken the time to learn about your stack, and build a buffer into your budget- Use the max possible draw from each AP and leave a cushion so you don't get blindsided when the CTO and all of his expensive friends come in for that holiday party, lighting up your APs with all the toys you wish you had. Below is the example from Meraki, one of the easier to read ones- Bonus points if you can tell what's wrong here, and why.

Power Budget (Right)

Power Use Sample


Next time, We'll get more properly technical, and dive into some common and uncommon Layer two issues- It won't be as brief, but it'll be more interesting.

52 views0 comments
bottom of page