No problem, take your time :)

Just now back to my desk at work. Managed to isolate the problem to a failing daughter card on one of the routing cores - was making a 10g fibre link which was 1/2 of a 20g port channel twitchy. The kicker... not ONE error in any log, anywhere. That was a pain to find. :/

My week has been reidiclous too. I have had to do so much extra work. Thursday was my only free day. That’s why I couldn’t get on to edit last night at my usual time :/ Was told will be getting a new guy on tuesday so should be back to normal this week and can edit. Just let me know when you add the links.

no worries take your time

Write erase....reload?

Oh, if only. The problem ended up being something we had to find by basically determining where the problem wasn't. After much narrowing down, it came down to two core devices and their link. After more testing, one of the 10g fibre channels that make up the bonded 20g port-channel was messed up. The module must be dying on one side, but producing no errors, log entries or anything that would highlight a problem. Only by shutting down one of the pair were we able to see the issue go away, so, temporary fix for now was to move stuff off the dying module to another free one and Monday we'll look at replacing it or the whole blade.

Ugh! Those are the worst.

Right? Well, I'm in this morning an hour early to go back over there and replace the daughter-card to see if that fixes it. If not, then it's reseat blade 3. If not that, replace module 3. >_<

Never underestimate the value of a little percusive maintenance. Give it a swat for me, eh?

Haha - and that's exactly what it was too - re-seat, clunk, all-good.