Hp Procuvre 2530 icmp packet lost ...

[lorenzo69]

Hi,

I have HP Procurve 2530 Switch (J9782A).

 I have a lot of (Request timeout for icmp_seq) when I ping my Switch.  but it isstill available with ssh or telnet.

after few seconds the ping become normal... and after few second again ICMP timeout.

But  I dont have problem with my server which is connected on this switch. I mean there is no problem when I ping my server when my switch is not pingable. (strange!)

I upgrade the firmeware of my switch, but not solved problem.

everyone know the cause of this problem, to give me the solution ?

Thanks

[Richard Brodie_1]

ICMP packets are handled by the CPU, rather than forwarded in hardware; and they get just about the lowest priority. So:

1. Have a look at CPU utilization when the switch starts dropping packets.
2. Don't worry too much about it.

[Chrisd131313]

Here'sa link to a HP advisory doc relating to switch ICMP responses...

http://h20565.www2.hp.com/portal/site/hpsc/template.PAGE/public/kb/docDisplay/?sp4ts.oid=51081&spf_p.tpst=kbDocDisplay&spf_p.prp_kbDocDisplay=wsrp-navigationalState%3DdocId%253Demr_na-c03251146-1%257CdocLocale%253D%257CcalledBy%253D&javax.portlet.begCacheTok=com.vignette.cachetoken&javax.portlet.endCacheTok=com.vignette.cachetoken

As Richard stated, it's likely your CPU utilisation is high when you experience dropped ICMP packets to the switch.

[lorenzo69]

Ok, but It make problème with my monitoring :( ...

[Richard Brodie_1]

If you're using ICMP as a health check, you could swap it for an SNMP probe.

Otherwise, it might be worth investigating if there is something that is hitting the CPU hard. I have several 2530s and don't generally see any ICMP loss.