CAT4500 stopped switching

Unanswered Question
Mar 3rd, 2010
User Badges:

Hi,


one of our Catalyst 4500 stopped switching and wasn't reachable until it was rebootet manually via console. There was no error message only the following message was logged permanently,


001417: .Mar  1 00:48:28.859 MEZ: Current Freelist count 511. Fell below threshold 580 times consecutively
001418: .Mar  1 00:48:58.906 MEZ: Current Freelist count 511. Fell below threshold 1181 times consecutively
001419: .Mar  1 00:49:28.957 MEZ: Current Freelist count 511. Fell below threshold 580 times consecutively
001420: .Mar  1 00:49:59.004 MEZ: Current Freelist count 511. Fell below threshold 1181 times consecutively
001421: .Mar  1 00:50:29.055 MEZ: Current Freelist count 511. Fell below threshold 580 times consecutively
001422: .Mar  1 00:50:59.102 MEZ: Current Freelist count 511. Fell below threshold 1181 times consecutively


Since we have rebootet the switch it works properly.Unfortunately I couldn't find any hint at cisco.com regarding this log message and the unexpected behavior.


used IOS is cat4500-ipbasek9-mz.122-31.SGA8.bin


Is there anybody who can explain the message or also had this problem.


regards,

Rocco

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Loading.
Sergei Vasilenko Wed, 03/03/2010 - 21:09
User Badges:
  • Cisco Employee,


/* Style Definitions */ table.MsoNormalTable {mso-style-name:"Table Normal"; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-parent:""; mso-padding-alt:0cm 5.4pt 0cm 5.4pt; mso-para-margin:0cm; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:10.0pt; font-family:"Times New Roman"; mso-ansi-language:#0400; mso-fareast-language:#0400; mso-bidi-language:#0400;}
Hello Rocco,

This could be an oversubscription situation due to broadcasts (storm, L2 loop?) or
due to traffic flooding from several ingress ports directed
to lower-rate ports (Gig->FastEth for example).
Like some servers/hosts sudenly start putting excessive traffic on the network.
You should see a lots of output drops on the interfaces connected 
to those oversubscribed ports as well, when it happening.
Messages looks like kind of indication of out-of-buffers conditions.

If it would happens again, in order to find from behind which ports flooding commes from
and which are oversubsrcibed make sense to check several time stats of
"show int", "show int switching" and possible "show spanning-tree detail".

Thanks,
Sergey 

Actions

This Discussion

Related Content