H.10.92 - ping loss and port-utilization error

[Kurns]

I used to be the one answering the questions, not asking them!

Since upgrading to H.10.92 I've noticed I get random bursts (every few hours on average) of sporadic ping loss when pinging my 2600's. I'm using PRTG which pings once every 30 seconds.

Previously on H.10.83 this wasn't an issue. I've tried using the management-vlan command, but that hasn't helped. The only other clue I may have is that there are about 1000 broadcast packets/second on this network.

I'm wondering what the fixes were between 83 and 92, as I'm sure one of them is causing this behaviour.

Another issue (only cosmetic) I've noticed is that 'show int port-utilization' can show very high values that are not inline with what is actually occurring on the port. Has anyone else seen this?

[Mohammed Faiz]

Hi,

Is H.10.92 an official release ? I hope not as I've just upgraded all our switches to H.10.83 as that's the latest listed on the site!

As for the port-utilization, what are you comparing the show values to? Is it the send/recieve values that look high or the utilisation percentage (which I understood as being an average over a few minutes)?

[Lynn-Marie]

I can get you a draft version of a fix list. :smileyhappy: Please make sure you report your findings to someone in support so that they can be addressed appropriately. I don't know whether this behavior has been seen by others or not.

[Lynn-Marie]

H.10.92 is released to support - they can distribute it to customers that need fixes, and it is fully supported even though it is not planned (to the best of my knowledge) for release to the web.

[Helper]

Hi,

Do you have any news ?

Is it a bug ?

Does Procuve plan to provide a new version for this product line ? What are the enhancement/bugfix since H.10.83 GA firmware ?

Thanks in advance

[Bruce Campbell_3]

Not necessarily relevant to your issue, but H.10.95 is available through support.  We reported a crash

after scp, and received this version (but we haven't tried it yet)

[Bruce Campbell_3]

We have approximately 550 of 2650's and 2626's at our site, of which

about 300 are on H.10.83 or earlier, and 200 on H.10.95 or H.10.101

We ping all these devices every 3 minutes, and randomly see loss of ping

(and confirm loss of ssh/snmp during these times) on the H.10.95 and

H.10.101 versions.  We have a case open.  Devices recover on their

own, from 10 minutes to hours, later.

H.10.83 or earlier is fine.