Switching Your Enhance Backup Server from BTRFS to Ext4 – The Easy Way: Page 2 - Enhance Control Panel

Switching Your Enhance Backup Server from BTRFS to Ext4 – The Easy Way

Andrei

wenani that's what I thought, but wanted to be sure. thanks.

cPFence

Andrei What do you mean by "Enhance getting confused"? If I just use the move website feature in the panel to move backups to a new (ext4) server will it not work?

We haven’t tested moving the backup role along with old backup data from BTRFS to EXT4, so I can’t comment on that until it’s been tested.

What we did was simply rename the backups on BTRFS, once you do that, the backups will disappear from your Enhance control panel, allowing you to move the backup role without any existing backup data.

I’ve seen some folks in the community reporting that BTRFS works well for them on the latest version, so it’s worth trying for a few days and seeing how it goes, especially if you only have a small amount of data.

Andrei

cPFence I've just tested moving the backup role along with backup data. A 300 MB website occupies 20 GB, and a 14 GB website occupies 400 GB on the new server. Seems excessive, I've investigate further.

cPFence

Andrei

That’s expected. BTRFS to EXT4 migrations are a pain in the neck. This process might take a day or two to complete!

I’m not sure of a completely safe way to stop it. You could kill the process, but that hasn’t been tested, so do it at your own risk.

rdbf

Andrei Seems excessive, I've investigate further.

Why not consider keeping the old backup setup? All that will happen is that the old snapshots will get deleted as they cross the retention horizon, while new backups will use the hardlink system.

Alternatively, if you need to switch servers, also set it up with a separate Btrfs block device and do a proper btrfs-send to just duplicate the current disk, and continue as above.

If switching to ext4 now means losing backup history or temporarily 50x-ing your storage and burning bandwidth, it's worth considering just waiting a bit to see how it works out with others.

Andrei

BTW, is there a way to stop the process of moving the backup role?

Andrei

FTP doesn't seem to work, either. I was close to giving this update to v12 a 10/10, but I guess there's a lot to be tested as issues keep creeping up.

cPFence

Andrei

Moving large backups from one server to another has always been a big task and would occasionally fail even in v11. You should have followed the guide outlined above for a smoother process and peace of mind.

Andrei

rdbf I'm keeping the old backup setup, I just want to test different scenarios.

twest

xyzulu welp of course I was one of the people running out of space on my v12 update, lol, I thought a spare 1TB would suffice, but it got to 99% usage so I just ran your ext4 conversion. Went real smooth, quick. Hopefully the fresh backups will play nice. This was such an elegant way of converting, flush away my old snapshots and bring on the new way 🙂

pratik_asabe

twest we also switched to ext4, so far so good.... it just that backup timing can go here n there but it seems stable than btrfs, btrfs on v12 was a mess so we decided to switch...

8Dweb

twest We had this problem with our backup servers, so we planned and installed a new backup server as part of the upgrade... but we still got stuck and had to unmount the /backups BTFRS partition on the backup servers and then decomission/delete them. So, we have the historical data preserved on those old servers, but started "fresh" with the new backup servers as part of the upgrade.

We also just ran into a weird "oops" where one client had backups on their own physical server (We do not recall why) and that came back to bite us yesterday since that backup role was missed in our v.12 upgrade. So, again, unmounted the partition, moved backup role to the EXT4 server and back in business.

Imagine all these small idiosyncratic things we each do/setup/utilize multiplied across many of us... no wonder there were so many revisions of v.12 in the first couple of weeks...

have you recovered now?

cPFence

pratik_asabe it just that backup timing can go here n there

Yeah, this bug showed up in the last two releases. It’s already been reported to the Enhance team, so hopefully, it’ll be fixed soon. Right now, the "Allowed Backup Hours" setting is being completely ignored!.

twest

8Dweb Yep just checking backups today and it looks great, usage is right on point for the amount of websites data we have, seems the conversion worked perfect.

I have another cluster to update soon, I'm just gonna plan to convert the backups storage as part of the process 😃

pratik_asabe

cPFence yes indeed! we noticed this when we decided to reduce the backup hours and scheduling only for overnight/non-production hour but backups didn't go through of course, so we enabled all backup hours but that didn't work either as backups are happening at very random hour! and as per Adam backups are rotated by hours not by quantity (bit confusing or proly i misunderstood idk) but by this logic retantion might also be cooked idk.. will notice it by tomorrow....

hope enhance team fixed this soon!

« Previous Page Next Page »