Cisco 2960L-24TS-LL - Management Issues

[francesco.rotorato]

Hello,

I'm experiencing an issues on my Cisco 2960L Switches...

As you can see from the screenshot, in the management network, there are different types of switches.

All of these have an IP on the management Subnet 10.20.95.X/26

All the Switches that aren't 2960L model, work fine, and are reachable in telnet/ssh.

Only the 2960L are unable to establish a telnet session, and they also loss 3 out of 5 ICMP packet if i try, from the Core switch (or from the other Access), to ping them.

I'm trying to find out what is happening to these machine, but I'm unable to find anything.

The show processes cpu, doesn't give me anything relevant...

And Cisco doesn't declares any bug (with this symptoms) affecting the running release.

I hope you can help me, I've also attached the show tech-support output.

Best Regards,

Francesco.

[DaSt-1986]

Are the default gateways configured correctly on the L-switches?

[francesco.rotorato]

---

@DaSt-1986 wrote:

Are the default gateways configured correctly on the L-switches?

---

I tried it, but it's always not working..

Thanks for your answer (Y)

[Mark Malone]

Hi

I ran your show tech through the cli analyzer it came back clean only informational options , no warning of hardware or software issues

I would still try another image on one of the L models to rule out any unknown bugs you can always flip it back , that would be my next step , not all bugs are listed in caveats section only the ones Cisco are aware about that have been triggered and identified

interfaces memory cpu all looks ok switch , there is a newer image there recommending might  be worth a shot 15.2.6E

don't use telnet either try stick to ssh , telnet is unsecure , I don't see any ssh configured but you have a K9 image so you can add it

[francesco.rotorato]

If I connect through console,
the switch, doesn't loss ping on itself.

The upgrade was my last chance...
I wanted to find something that could help me to understand,
why this is appening.

Thanks for your Reply, if I won't find anything else, I will upgrade.

[Mark Malone]

pinging switch wont give a valid reply as the icmp is not controlled in copp , so if there's any load on the switch during the ping it can fluctuate , a true icmp test is end to end through the switch pc - pc or use ip sla ,but if its dropping your actual terminal session through SSH there is an issue , but pinging switch is not a valid test for sure on any platform unless its controlled icmp

[francesco.rotorato]

PS: I tried to configure ssh, but it's slow and unstable as the telnet session at the moment...

[Mark Malone]

Just so we know where your coming from how are you seeing it as slow , is it the terminal itself when you type commands it takes ages to appear on the screen or what way are you seeing it that its slow compared to other non effected switches ?

[francesco.rotorato]

It's slow compared to other non affected switches,
and it also takes ages to appear on the screen :P

[Mark Malone]

Are you centrally accessing all of these switches form same terminal or does it be from different laptops , just something you could check before upgrading if always same machine try another one as a test and try a few different terminals like , putty , hyper v , secure crt , tera term , maybe the L model doesn't like the software terminal connection , to rule it out

[francesco.rotorato]

Tried out a lot of different ways, to connect.
Direct OS X / Windows terminal, Putty, Secure CRT...
From the PC, from the Core, from an other non-L switch...
From remote VPN, from LAN...
And all the L-Series, doesn't establish a good session.
But a week ago they was working fine...
After the installation of a Cisco WLC-2504 they started to not answer good.
But I don't think this is the cause of the fault.

[Mark Malone]

hmm seems coincidental , I know its not a normal working hours thing but if you revert to the old design is it working fine as a test even disable it for a few minutes , seems odd a new device gets introduced and now some network devices are gone slow that were working, other than mgmt. connection though the switches are working fine yes , processing traffic no slowness for the users that's directly connected to them , looking at the doc there doesn't seem to be a dedicated MGMT port which might have been a way to avoid this