We have a native stack with 4 SG350X switches and testet failure behavior (take members from power). A failure of member 3 and 4 didn't hade a big influence, sometimes 1 ping lost. But when you unplug the slave all devices on the stack are not reachable for about 15-20s. Same behavior it comes back (plug in power again). If you unplug the master same behavior, also all devices loose network connection for about 15-20s. And what irritate us absolutly is when we plug in master again. First again this lost connection for about 15-20s but then it reboots the slave and it looks like ths stacks do a reboot, b'c it takes about 3min until all devices have network again. Last test was to unplug all switchs at once and put them back on power in random order. That also tooks about 3min and in this case this is of course ok, but not the other observed lost of connections.
Can someone give a bid light into the darkness!? We think that this is not the expected behavior, a few seconds are ok, but 3min is definitly too long and we also don't see any reason why rebooting a slave when master is comming back online
