Tuesday, March 2, 2021

Slow network and ping loss

Hi,

so we added a 4th switch to our core stack (ring) and a new SFP LAG uplink to a new access switch stack.
Also upgraded all stacks to the latest 2.5.5.47 firmware and everything seemed fine on sunday.
Models are all Cisco SG350x in different flavors (with/without POE, 48P, 24P).

With the returning users on monday, shit started to hit the fan.
Users are reporting problems on time critical applications (like Microsoft Dynamics nav).

Tests show that ping response times sometimes jump from <1ms to 2ms, sometimes 15ms, or drop completely. I don't see STP blockings on any stack.

CPU Load on the core stack jumps up and down from 90% to 40%.
Cisco Webinterfaces are extremly slow too.

I'm kinda lost here, been wiresharking the network for some time, i don't see anything too freaky.
I'm wondering if we simply hit the performance threshold of the SG350x (about 300 endpoints) or if the new firmware did something unexpected, although i didn't see much in the release notes, except changes in findIT.

Any help or advice for further debugging would be greatly appreciated.



No comments:

Post a Comment