We had a problem with our 9.211 systems that for one reason or another the slave would stick in the UP2DATE process and our HA would then become non-resilient.
It seemed whenever we'd have an update it would get stuck in UP2DATE until support logged in and freed the process somehow.
So as we needed to migrate to 9.3 to make use of some new features we needed first to make sure that the current pair was in sync. Try as we might the slave just remained in UP2DATE status.
So after asking support we were asked to reimage the slave. Which seemed easy enough, but the only ISO for 9.2 is 9.214, which has no upgrade path to 9.3 - yet [:(]
The only thing to do was bite the bullet, break HA and upgrade the now standalone box to match the latest available 9.3 ISO which was 9.314. This way we could upgrade them both to the same release.
Did that and then tried to enable HA.
No joy at all the slave now a factory default unit doesn't seem to activate any ports at all - including eth3 which is used to auto configure HA. So no way to add the slave into the HA automatically as we did back in the day when we installed from scratch in 9.2.
The way forward was to backup the config from the standalone. Download it onto a Linux (ext2) formatted USB stick. Mount the USB stick at the command line in the slave and restore the backup file using backup.plx
To me this meant we'd end up with two standalone boxes and all sorts of ARP issues, but it seemed to work. Logged onto the GUI and turned on HA and after quite a few bankend failures things started to SYNC. We had some service disruption for about 5 minutes, but things are now back as they should be.
So what happened to the automatic HA config?
This thread was automatically locked due to age.
Guest User!
You are not Sophos Staff.