01-21-2015 08:29 AM
Here is my scenario that we are troubleshooting.
We are in the process of decomissing several arrays. When offlining the switchports we notice that we get C3 discards on ISL's between the switch that the port was offlined and its peripheral ISL switch. Its not a lot, i would say it drops a frame every 5-10 seconds.
Whats strange in this scenario is that there is no zoning on any of the ports, the arrays are not connected to any servers for a while ( more than 2 weeks).
1) offline an switch port
2) notice c3 discards on ISL
3) would enable the switchport back on again and the C3's would stop.
Has anyone seen anything like it in their environment ? My san is big, many switches, many servers connected so its hard for us to trace out what is causing it.
We have logged a support case with brocade and they said that it may be a server that has some reminants of the array nport that it was once logged into. We have verified it by looking at the server and seeing that this server was once connected to an array port. Question now becomes do we have to look at every server that were once on this array... is there any other way to go about this? Has anyone ever seen this in their environment ?
Solved! Go to Solution.
01-21-2015 08:54 AM
what FOS version do you running? what type of discards you getting
you can check framelog where you will find TOP 20 discards check the reason SID DID then from 7.2.x you can monitor flow .
01-21-2015 09:20 AM
not going to happen... why do I even bother asking support forums... its like calling support and talking to a L1 guy who has no clue of whats going on....... SMH.
01-21-2015 03:24 PM
01-23-2015 12:23 PM
Alexey, you are exactly on the money as they say.
In your case it was AIX in our case its all legacy HPUX.
portlogdump was exactly what we did to identify strange activity on the port that was believed to be causing the C3 discards.