[Plugin] CA Appdata Backup / Restore - Deprecated

[SelfSD]

8 hours ago, Squid said:

The backup finished.  But then it went through and was trying to delete the old dated backup sets.  This is where it was when you grabbed the diagnostics:

 

rm -rf /mnt/user/backups/appdata_backup/[email protected]

 

 

Assuming that this backup set was ~200Gig all small files, this will take a while to complete.  However, the command itself will not impact anything else in unRaid.

 

There is nothing that I see in the diagnostics as to why your system was running at a snail's crawl.  (And your CPU is better than mine)

 

One or two users have issues with the standard bash commands used by the backup script (rsync & rm).  No specific cause has been found, and any issue with those commands is completely out of my control.

 

The best I can offer is to install the tips & tweaks plugin and try this:

 

 

 

 

I've applied the new values. I'll let you know how it goes next week.

[tucansam]

Gang,

 

I have my 'appdata' folder split between two SSDs, /mnt/disks/ssdplex/appdata is dedicated to plex, and /mnt/disks/ssddocker/appdata is dedicated to everything else.  Two separate SSDs.  

 

I also have a separate spinner for sab downloading and unpacking, /mnt/disks/wd500

 

Can I still use this plugin to backup the two SSDs, and perhaps even the wd500?

[Squid]

12 minutes ago, tucansam said:

Gang,

 

I have my 'appdata' folder split between two SSDs, /mnt/disks/ssdplex/appdata is dedicated to plex, and /mnt/disks/ssddocker/appdata is dedicated to everything else.  Two separate SSDs.  

 

I also have a separate spinner for sab downloading and unpacking, /mnt/disks/wd500

 

Can I still use this plugin to backup the two SSDs, and perhaps even the wd500?

Nope.  Only supports a single appdata folder.

[SelfSD]

My server froze again tonight but I got nothing to show for it this time. No response from SMB or the WebGUI when I woke up like last time. There where no errors or warnings on the connected display either. I was able to reach the password prompt via SSH but nothing happened past that.

 

Before I went to bed last night I enabled Troubleshooting Mode in Fix Common Problems but it hasn't saved anything to the USB. The write counter did go up for the few minutes I watched it, thinking it was saving the log to the flash drive but there's nothing in there but old diagnostics.

[shooga]

@SelfSD I've had this same issue. Not really a fix, but I have found that (so far) the problem is avoided by adjusting my schedules so that Mover runs at midnight and this backup runs at 4am. That seems to leave enough time for Mover to finish so there isn't a conflict. I wish there was a more concrete solution, but so far so good...

[SelfSD]

3 hours ago, shooga said:

@SelfSD I've had this same issue. Not really a fix, but I have found that (so far) the problem is avoided by adjusting my schedules so that Mover runs at midnight and this backup runs at 4am. That seems to leave enough time for Mover to finish so there isn't a conflict. I wish there was a more concrete solution, but so far so good...

 

So it's a conflict with the mover? I'll set it to a few hours before the backup schedule and see how that works out next week. Thank you!

[shooga]

6 hours ago, SelfSD said:

 

So it's a conflict with the mover?

It's just a theory really, but I couldn't think of any other difference between a manual run of this backup script and the scheduled ones. I never had an issue with a manual run. This work around seems to be helping so far.

 

Also: It's not really clear above, but I run mover every night and the backup script once per week. I wanted to make sure there was no overlap on the day they both run.

[alfi33]

16 hours ago, SelfSD said:

My server froze again tonight but I got nothing to show for it this time. No response from SMB or the WebGUI when I woke up like last time. There where no errors or warnings on the connected display either. I was able to reach the password prompt via SSH but nothing happened past that.

 

I've also been having this issue for several weeks.  The last time it froze I pinged the server and got replies. I disconnected the NIC and reconnected it and the pings stopped and did not resume...which I found a little odd.  I have to hit the power to get the server back up.  In addition, I've got my server joined to an AD domain and when it comes backup after hitting the power it is unjoined.  

 

I've been able to do a couple manual backups without issue.  But I don't think my problem is due to overlapping with the Mover schedule.  My Mover and Backup schedules are 3 hours apart and Mover completes in well under an hour.  

 

I've been wondering if it's a particular Docker that's causing the issue because I might not have had them all running when I ran Backup manually.

Edited July 16, 2017 by alfi33

[alfi33]

3 hours ago, alfi33 said:

I've been able to do a couple manual backups without issue.  But I don't think my problem is due to overlapping with the Mover schedule.  My Mover and Backup schedules are 3 hours apart and Mover completes in well under an hour.  

 

I've been wondering if it's a particular Docker that's causing the issue because I might not have had them all running when I ran Backup manually.

 

Well, I fired up all my Dockers and kicked off a manual Backup and my server froze just as it has been doing for scheduled backups.  I had Troubleshooting Mode running this time so I've attached what it captured.  Also attached a picture of the local monitor when it was frozen.

 

I was able to start a tail from the console but the console became unresponsive after that.  CTL+C did nothing.

The server was still pinging.  Unplugged the NIC and plugged it back in.  Pinging stopped even after reconnecting NIC.

 

Next, I might try another backup without some of the Dockers running....but doing a parity check from the unclean shutdown each time slows down the troubleshooting.

tower-diagnostics-20170716-1410.zip

FCPsyslog_tail.txt

[Squid]

You (and 2-3 other users) all seem to have the same problem in the same part of this.  Namely where its deleting the dated backup set that's too old to keep.

 

This is done by the rm bash command (in your case)

rm -rf /mnt/disk3/CommunityApplicationsAppdataBackup/[email protected]

 

This is out of my hands, and I have no clue as to why its failing.

 

But, ultimately, the rm command doesn't really even need to be executed if you reconfigure the plugin a bit.

 

Either forgo dated backup sets, and just use a single one.  Or, enable attempt faster rsync, and make sure that your backup frequency and the time to keep dated backup sets means that there is only ever a single set that will ever get deleted when a backup runs.  The faster rsync option instead of deleting the dated backup set, instead renames it to what the new set would be called and then rsyncs into it, only copying files that have changed.  Running backups manually however will result in there being an extra backup set at some point being present that will result in a deletion.

 

Personally, with my use case, I don't see why I need a backup from this week, and the one from the week before, so I don't use dated sets.

[alfi33]

Thanks for looking at this @Squid.

 

I tried manually deleting some of the older dated backup folders (using both Dolphin and Windows Explorer).  I could delete most of the contents but I got errors on some of the files.  Looks like permission might have gotten screwed up on some of them.   

 

 

 

I tried modifying the permissions using Dolphin but my changes didn't seem to stick (not sure if I'm doing it right though).

[alfi33]

17 minutes ago, alfi33 said:

 Looks like permission might have gotten screwed up on some of them.   

 

Well, I can delete the folders just fine with Midnight Commander (Still trying to get the hang of Unix permissions).

 

Once I get all these old folders cleaned up, I'll try another backup (with faster rsync enabled) that will require only 1 dated folder to be deleted.  We'll see how it goes.

[SelfSD]

14 hours ago, Squid said:

Either forgo dated backup sets, and just use a single one.  Or, enable attempt faster rsync, and make sure that your backup frequency and the time to keep dated backup sets means that there is only ever a single set that will ever get deleted when a backup runs.  The faster rsync option instead of deleting the dated backup set, instead renames it to what the new set would be called and then rsyncs into it, only copying files that have changed.  Running backups manually however will result in there being an extra backup set at some point being present that will result in a deletion.

 

Does this mean that if I want to keep two backups, one per week I should set the "Delete backups after how many days?" to under 14 days?

 

I would like to have one untouched backup in case the server crashes during a backup

[showstopper]

I have been trying to look into backup solutions for unraid - this looks very nice.

So does this do vm backups and can you restore files from the vm backup?

 

Essentially i am looking to build an unraid server that would have a plex docker and my media (that i would like backed up).

I would also be building a small home domain, file and Exchange vm's - would this product be able to back all these up individually so i could restore files from the file server vm, and also the entire vm if required??

 

Trying not to ask too many questions (sorry!) does this, or any other app, allow for vm snapshotting?

 

Cheers.

[bobbintb]

3 hours ago, showstopper said:

I have been trying to look into backup solutions for unraid - this looks very nice.

So does this do vm backups and can you restore files from the vm backup?

 

Essentially i am looking to build an unraid server that would have a plex docker and my media (that i would like backed up).

I would also be building a small home domain, file and Exchange vm's - would this product be able to back all these up individually so i could restore files from the file server vm, and also the entire vm if required??

 

Trying not to ask too many questions (sorry!) does this, or any other app, allow for vm snapshotting?

 

Cheers.

This will back up your VM configuration. Any files on the drive image of the VM will have to be backed up and treated as a separate machine.

[showstopper]

This will back up your VM configuration. Any files on the drive image of the VM will have to be backed up and treated as a separate machine.

Ah ok thanks for this

Sent from my SM-G925F using Tapatalk

[Squid]

On 7/17/2017 at 11:23 AM, SelfSD said:

 

Does this mean that if I want to keep two backups, one per week I should set the "Delete backups after how many days?" to under 14 days?

 

I would like to have one untouched backup in case the server crashes during a backup

Backup frequency: weekly (7 days)

Days to keep backups will be 9-13 days

 

Day 1: Backup Performed (New)

Day 8: Backup Performed (New)

Day 15: Backup Performed (Overwrites Day 1)

.

.

.

Manually performing a backup on say day 3 will result in there being 2 backups on Day 15 being scheduled for deletion.  Once will get overwritten, but the other will get deleted at which point the linux rm command will puke.  I can probably work around the rm, but the process will be very noticeably slower

 

[SelfSD]

5 minutes ago, Squid said:

Backup frequency: weekly (7 days)

Days to keep backups will be 9-13 days

 

Thank you! I have manually removed the older backups and only kept the two latest. I'll report back on Sunday with how it went with Attempt faster rsync enabled.

[Squid]

6 minutes ago, SelfSD said:

 

Thank you! I have manually removed the older backups and only kept the two latest. I'll report back on Sunday with how it went with Attempt faster rsync enabled.

If nothing else, the backups after a week or two will only take ~10 minutes instead of 1 hour as it only has to copy changed files

[NoobSpy]

@Squid. Is it possible to extend this script to include Vm images as well?

[alfi33]

On July 16, 2017 at 8:19 PM, alfi33 said:

 

Once I get all these old folders cleaned up, I'll try another backup (with faster rsync enabled) that will require only 1 dated folder to be deleted.  We'll see how it goes.

 

No freeze or crash this week.  And the backup only took 4 minutes.  Thanks for your help, @Squid.

 

I wonder if my problem was just that I had more than one dated backup folder that needed to be deleted...or if it was something about a particular dated backup folder that the backup routine was getting hung up on?

[Squid]

13 hours ago, NoobSpy said:

@Squid. Is it possible to extend this script to include Vm images as well?

I incorporated the ability to run user scripts within it.  You can use the auto_vm_backup script from @danioj  to accomplish this.

 

@jonp has previously stated (IIRC) that snapshots, etc would get incorporated soonTM into unRaid so I've been hesitant on adding the feature into the plugin.  But, should my VM ever die on me, then I'll probably incorporate it as I'd be kicking myself at that point.

[SelfSD]

10 hours ago, alfi33 said:

 

No freeze or crash this week.  And the backup only took 4 minutes.  Thanks for your help, @Squid.

 

I wonder if my problem was just that I had more than one dated backup folder that needed to be deleted...or if it was something about a particular dated backup folder that the backup routine was getting hung up on?

 

Everything looks good so far as well! My backup took 16 minutes vs 1h30!

[Squid]

10 hours ago, alfi33 said:

I wonder if my problem was just that I had more than one dated backup folder that needed to be deleted...

Definitely not that.  During development, I was regularly testing have 20+ backup sets that needed to be deleted.

4 minutes ago, SelfSD said:

or if it was something about a particular dated backup folder that the backup routine was getting hung up on?

Its the linux kernel (specifically the remove (rm) command) that's getting hung up.  No clue as to why it happens for a very small subset of users.

[NoobSpy]

10 hours ago, Squid said:

I incorporated the ability to run user scripts within it.  You can use the auto_vm_backup script from @danioj  to accomplish this.

 

@jonp has previously stated (IIRC) that snapshots, etc would get incorporated soonTM into unRaid so I've been hesitant on adding the feature into the plugin.  But, should my VM ever die on me, then I'll probably incorporate it as I'd be kicking myself at that point.

@Squid Cheers

Edited July 23, 2017 by NoobSpy