We're working to pinpoint the cause of a stack master re-election.
Here's the scenario:
Setting all ports on a 3750 48 port switch to vlan dynamic, using the int range command, 24 ports at a time.
Enter configuration commands, one per line. End with CNTL/Z.
OTCSTACK5(config)#int range Fa2/0/1 - 24
OTCSTACK5(config-if-range)#switch acc vlan dynamic
Jun 22 20:12:31: VQPC PAK:
Jun 22 20:12:31: VQPC PAK: HRPC HANDLER rcvd 74 bytes from box 1:
Jun 22 20:12:31: VQPC: rcvd HRPC_REQ_VQP_QUERY from box 1
Jun 22 20:12:31: VQPC PAK: xmt transaction ID = 0x000000B1
Jun 22 20:12:31: VQPC PAK: sending query to VMPS
Jun 22 20:12:31: VQPC PAK:
Jun 22 20:12:31: VQPC PAK: rcvd packet from VMPS
Jun 22 20:12:31: VQPC PAK: transaction ID = 0x000000B1
Jun 22 20:12:31: VQPC: Master sends HRPC_REQ_VQP_RESPONSE to slave 1
Jun 22 20:12:31: VQPC PAK:
Jun 22 20:12:31: VQPC PAK: HRPC HANDLER rcvd 16 bytes from box 1:
Jun 22 20:12:31: VQPC: rcvd HRPC_REQ
Read from remote host 10.0.0.165: Connection reset by peer
Connection to 10.0.0.165 closed.
30 seconds later we are able to reconnect and get:
OTCSTACK5#conf t
Enter configuration commands, one per line. End with CNTL/Z.
OTCSTACK5(config)#int range f2/0/1 - 24
Command suspended while switch(es) are being added to stack...
Command suspended while switch(es) are being added to stack...
OTCSTACK5(config-if-range)#end
a quick show switch shows
Switch# Role Mac Address Priority Version State
----------------------------------------------------------
*1 Master 001b.2b11.5300 1 0 Ready
2 Member 001b.2b11.5380 1 0 Ready
3 Member 001a.e37d.6a80 1 0 Ready
4 Member 001d.4654.6880 1 0 Ready
This is the prior stack master election
Switch# Role Mac Address Priority Version State
----------------------------------------------------------
1 Member 001b.2b11.5300 1 0 Ready
2 Member 001b.2b11.5380 1 0 Ready
*3 Master 001a.e37d.6a80 1 0 Ready
4 Member 001d.4654.6880 1 0 Ready
Switch 3 was the stack master, switch 1 is now the current.
No power loss, no reboots.
It looks to me like the most probable cause would be a cpu load related re-election, which is somewhat concerning.
Has anyone experienced anything similar to this? Anyone able to shed any additional light on it?
Any assistance would be appreciated.
