How do i delete backups manually?

slimx

With the current backup issues, what is the best way to actually delete old backups? Since we can't delete it from enhance itself.

I don't want to go in one by one, surely there's a command that's more efficient?
Server officially filled up. 🙁

cPFence

slimx

Here’s the code we use on our test cluster to clear old backups quickly. Just change the days variable to your needs and copy-paste it into the terminal (proceed with caution and always run the dry run version first):

https://gist.github.com/cPFence/12ea501396cb9df74100b61a9ba497b7

Out of curiosity, what issues are you running into with backups? Did you open a ticket with Adam? It's a bit strange you're still having problems, as we haven’t seen any backup bugs on any of the clusters we have access to at the moment.

rdbf

slimx

Simplest way is just delete a few of the snapshot folders, starting with the older ones off course. There's no database entries or anything, the control panel just reads the folders on the backup folder and that's it.

I can confirm the de-duplication issue has been fixed for all websites since v12.0.24. But it requires a command to run, which says that this should only be done at the instruction of Enhance support. Just ask Adam in a quick email, de-duplication will then kick in, and disk usage will go down with backups crossing the retention horizon.

Steamon

Its true that I did not managed to find a back-up delete button in gui is that coming?

cPFence

Steamon

Yes, it’s already on the roadmap.

slimx

cPFence
Backups are not functioning as incremental.
Storage space fills up very quick because of this. (2-3 days) - A few of the client websites are huge.
Enhance team says they are still working on a solution. Tickets are open.

I'm gonna give your script a test run now 🙂 hopefully it works

cPFence

slimx

That’s interesting. Keep us updated if you manage to figure out what the issue was. That would be super helpful for us, especially since you started fresh with EXT4.

rdbf

slimx Backups are not functioning as incremental.

Websites are not having app + db + email all on one server, but email separate?

Vendoz

rdbf i have same server for app-db and no email service in my cluster, and had same problems, so email servers aren’t the cause. I created a new backup server EXT4 and, NOW, it seems it works, but for the first days it did only full backups, even the new server. In addition, every backup role movement for every website create a x3/x4 disk usage on target server. With live monitor, i saw usage down from source (2gb) and up in the target (for 10gb), so i think that the problem was in the hard links management and coding process.

With 12.0.22 and the fix permission command i have to say that, for the moment, latest backups are incremental and restore works as expected.

But I have to wait few days cause the rotation should clean the old bugged full backups, I went from 250GB to 900GB actually 😃

slimx

rdbf

For 99% of the websites, APP + DB is on the same server
I only have a couple websites with a SEPARATE DB server.
For EMAIL 100% of websites are on DIFFERENT server.

Vendoz With 12.0.22 and the fix permission command i have to say that, for the moment, latest backups are incremental and restore works as expected.

Yes in my case it appears that 60% of the sites are working as expected, but a fair few are not working. RESTORES etc. are working fine, i tested it. It's just the back up sizes they're all full backups.

DracoBlue S3 copies are also not working properly; they do not follow the scheduled execution timing or the maximum backup retention.

I can confirm same issue, I disabled S3 for now as it was only an additional backup solution for me. Expensive here in OCEANIA anyway lol.

mendozal I had an issue with regular backups where the schedule was not being followed. I had a 24 hour minimum backup age but the server was taking backups every 12 hours. Adam told me to set and save the values again in /settings/service and it started to take backups every 24 hours as scheduled. Maybe this could apply to S3 too.

I had the same issue with regular backups and that recommendation WORKED

DracoBlue

S3 copies are also not working properly; they do not follow the scheduled execution timing or the maximum backup retention.

mendozal

DracoBlue I had an issue with regular backups where the schedule was not being followed. I had a 24 hour minimum backup age but the server was taking backups every 12 hours. Adam told me to set and save the values again in /settings/service and it started to take backups every 24 hours as scheduled. Maybe this could apply to S3 too.

DracoBlue

mendozal He advised me that too, but it didn't work.

cPFence

DracoBlue

Try incremental backups. There is no need to set up a new server to test it out. Just test it for a site or two by adding the backup role to any of your current servers.

SeLi

Hi you also have a script to check if backups exist? It seems after I was create a backup server and move all pages to that backup server it not create backups on him. (not happen on each page...)

xyzulu

SeLi Hi you also have a script to check if backups exist?

@cPFence provided a neat script for that here: https://community.enhance.com/d/2476-enhance-v12018-incremental-backups-now-working-as-expected/92

You can also check manually in the folder you chose for your backups.. website id's and snapshot folders along with a folder called "current" in each should exist if backups are working correctly.

rdbf

slimx For EMAIL 100% of websites are on DIFFERENT server.

For this, the v12.0.24 ecp CLI tool will fix the de-duplication issues if you ask Adam

slimx I only have a couple websites with a SEPARATE DB server.

This doesn't matter fortunately.