I therefore started the update procedure from the webgui. I did not get any error messages, but now it looks like my Master is 6.201 and my Slave is stuck at 6.200. I've looked a bit in the logs:
# cat up2date.log
2007:09:18-21:38:13 (none) ainstall[337]: System Up2Date: Started installation in Ha-Master-Mode upto version 6.201.tar.gpg
2007:09:18-21:38:31 (none) ainstall[337]: System Up2Date succeeded: Installed Up2Date Package Version 6.201
2007:09:18-21:38:31 (none) ainstall[337]: System Up2Date: New System Up2Dates installed
2007:09:18-21:38:33 (none) ainstall: HA_check: Ha Version conflict. Master: 6201 > Slave: 6200. Updating slave machine
2007:09:18-21:39:27 (none) ainstall: HA_check: Ha Version conflict. Master: 6201 > Slave: 6200. Updating slave machine
# cat high-availability.log
2007:09:18-21:33:20 (none) ha_master[2727]: Setting master version to 6.200
2007:09:18-21:33:25 (none) ha_master[6774]: UDP SEND sendto error
2007:09:18-21:38:31 (none) ha_master[2727]: Setting master version to 6.201
2007:09:18-21:38:32 (none) ha_master[2727]: Setting master version to 6.201
2007:09:18-21:38:34 (none) chk_lbeat[3719]: HA check: no link beat on interface - retrying
2007:09:18-21:38:36 (none) ha_master[6774]: UDP SEND sendto error
2007:09:18-21:38:57 (none) ha_master[2727]: Setting master version to 6.201
2007:09:18-21:39:01 (none) ha_master[6774]: UDP SEND sendto error
2007:09:18-21:39:02 (none) ha_master[6774]: UDP SEND sendto error
This is a production system so I can't just pull them down and "play around".
Do I need to deactivate the HA and manually update the Slave? How? Shutdown master and then update the slave (which will take over) and then power on the master again?
This thread was automatically locked due to age.