First, sorry for my English, as it's not my first language. Also, networking is not my specialty, we're a small enough business to go by with only one IT guy and I'm supposed to "know it all".
I'm having weird issues with our HQ local "core" switch. Every now and then some (random) parts of network traffic stops flowing. I'd understand this better if everything would just stop working, but no, some users (even in the same access switch) may be able to do work as usual, while some have nothing, no Internet, no access to servers. Some have reported to have access to Internet, but not servers. Whether they had connections open prior the issue begins or not doesn't seem to matter.
Our local network consist of operator CPE, which WAN is connected to MPLS (to our branch offices, Internet) and LAN to the problematic 2530. GW and routing are done by the operator. The eight access switches we have (more 2530 series ones, old ProCurves and one HP OfficeConnect), Hyper-V server with NIC Teaming and some misc. security devices are connected to the "core" 2530. There are 9 VLANs. The issue spans over different VLANs as not only workstations lose connectivity, but equipment in the other VLANs also goes down. This for example causes some of our doors to stop functioning, but again, not all of them. Access from MPLS to our servers is compromised as well.
All issues do go away for couple of weeks to even months by restarting the "core" switch.
We're currently running YA.16.02.0027 on the "core". Access switches have varying firmwares. Spanning Tree is turned on on all of the switches. There's nothing on the switch alert log, nor is there excessive amounts of errors logged.
We had similar issues two years ago and then I replaced the "core" switch with a new one and the issue seemed to go away, until this year. Do you guys have any idea what to look at and if I'm to replace the switch again, what switch would you recommend? I'd rather stick with HP as almost everything else we have is HP gear.
No comments:
Post a Comment