EXOS 16.2.2.4 / BD8810 : Unexpected slot reset ( MSM-A: pibConduitMasterRcvOneSlot(): failed, rc=-6, errno=134, slot=3)

[ATANGANA_NGA_Ar]

Hello, Please have some of you ever experienced this problem on Black Diamond 8810 : Unexpected slot reset while the switch was behaving normally before the failure. This happens from time to time and creates outage for a couple of minutes (restart of the slot)
05:49:04.62 MSM-A: Slot-3 FAILED (3) Conduit receive error encountered05:49:04.62 MSM-A: Slot-3 FAILED (3) Conduit receive error encountered
05:49:04.62 MSM-A: Slot-3 Error. Reason = Conduit receive error encountered (27) 05:49:04.62 MSM-A: pibConduitMasterRcvOneSlot(): failed, rc=-6, errno=134, slot=3.Please does anyone know how to fix ?

[EtherMAN]

We are running several 8900's in our network and have been since they were introduced. One question that comes to mind since you mentioned memory is what model cards and MSM are you running. Huge difference in the XL cards and the MSM 128 vs the C series when it comes to available memory and hardware resources. One thing you can do during an outage window is run extended diagnostics on the card you are having these issues with... Only the interfaces on that card will be affected and it will be about 5 to 10 minutes for the card to complete and come back online.

TAC would be able to steer you and give you more debug commands you can run to see if anything is going on...

[simon_bingham]

I believe a Conduit error is a backplane issue, someone from extreme can probably reply more knowledgably than me. if you have a support contract raise it with support.

[ATANGANA_NGA_Ar]

Dear Simon,

Thanks for your reply.

Yes I am thinking about raising a case to support... but I already checked Backplane and it seems there is no issue there. Can this be related to memory available on the slot because that slot has the least free memory among all. Or is it a known bug in 16.2.2.4 ? --> I would like to know if it is recurrent in the community so that I plan an upgrade

BR