Lost connection with unit 2 reason : connection timeout. Unit will be rebooted .

[henrihoffmann]

Hi,

have here a nativ 2*SG550XG-8F8T stack. Today I received this messages.

It looks like only unit 2 was rebooting, but the unit 1 master did not respond to pings, at least 1-2 minutes at this time. Therefore our HA disk appliance (iSCSI & NFS) died. When all ping nodes are down for a longer period of time, the HA solution is shutting down to avoid a split brain situation.

Henri

2147480967 2018-Jul-16 07:16:01 Error %Stack-E-STCK-EXCEP-HNDLR: Lost connection with unit 2 reason : connection timeout. Unit will be rebooted .
2147480990 2018-Jul-16 07:09:53 Error %Stack-E-STCK-EXCEP-HNDLR: Lost connection with unit 2 reason : connection timeout. Unit will be rebooted .
2147481031 2018-Jul-16 07:07:13 Error %Stack-E-STCK-EXCEP-HNDLR: Lost connection with unit 2 reason : connection timeout. Unit will be rebooted .

[ktonev]

Hi,

What model of SFPs are you using to connect the 2 switches?

[henrihoffmann]

Hi,

no SPFs, we use 2 CAT7 cables.

Henri

[henrihoffmann]

Hi,

anybody, who can explain, why the master unit does not repsond to any ICMP requests for more that 2 minutes, even just the backup had a problem? From the paper point of view, this should be a HA solution, at my knowledge, the stack is necessary for the Etherchannel configuration.

Thanks

Henri