j12t wrote: ↑Sun Oct 11, 2020 3:24 am
It depends on exactly where the problem is during upgrades. Depending on that, the restore after the update may or may not make it all the way through. We do try and keep data safe by not deleting anything unless it looks like everything worked out.
If there is a kernel update, or something like that, that requires a reboot:
1) Before the reboot backups are created, sites are undeployed and the packages are updated
2) Reboot
3) After the reboot, the sites are redeployed
We don't redeploy before the reboot, because the sites would be redeployed against the still-running old kernel and other old daemons.
Hope this clarifies.
Re mastodon, yes, this is one of the reasons why mastodon is still on the yellow channel and hasn't been promoted to the green channel. It's a rather "expensive" piece of software to keep updated, and given we currently have no revenue stream to support that development, it has not gotten as much attention as we'd like. Maybe one day we'll sell a UBOSbox mastodon, but not yet.
Thanks for your feedback.
1. Now I understood the process that UBOS takes when updates proceed.
But still I have a doubt, in my case, after the reboot it appears sites are redeployed as you explained, but at this point UBOS refuses to invoke
since, there are still files on the "/UBOS/backups/update folder", which causes "“Cannot create a temporary backup; the backup directory is not empty”" problem, as it is described in the trobule shooting section. Beforehand, I didn't see this message at all after the udate, so I have a question about this.
2. Yes, I do agree with your point, as a result UBOS box would not be able to work as we expeced after some updates. In this case, is it a problem on the Mastodon side? And do we have a chance to look at the issue and fix it? I personally think, just like UBOS, Mastodon is one of the best platform for the decentralised communications and looking forward to see a great collaboration with UBOS. Thanks.