Plugin Contributor
bcrodua
(@bcrodua)
Hi,
Apologies for the delay over the weekend.
To help us work out the cause of your problem, can you please send us a copy of the latest full backup log?This can be found in the UpdraftPlus Existing Backups tab. The contents will be too long to post here directly, but you can use an online service such as Dropbox or Pastebin, and post the link here.
Regards,
Bryle
Hi,
is there a way to provide the log w/o publicly posting a Dropbox or Pastebin URL?
Regards & thanks,
Tim
Plugin Contributor
bcrodua
(@bcrodua)
Hi Tim,
It is safe to share the log with us here. You can change the site URL in the log file if you need to.
Thanks,
Bryle
Hi Bryle,
perfect – you´ll find the log at https://pastebin.com/7PPQKWBY .
Thanks,
Tim
Plugin Contributor
bcrodua
(@bcrodua)
Hi Tim,
Thank you! Can you try to reduce the split archive size further down to 25mb? Then run a test backup again.
Thanks,
Bryle
Hi Bryle,
unfortunately, that didn’t cure the problem. As far as I can see, the logs pretty much look like before (the log simply ends during upload).
Thanks,
Tim
Plugin Contributor
bcrodua
(@bcrodua)
Hi Tim,
Can you try to ask your host to increase your server I/O limit, maximum_execution_time and site memory?
I believe it has something to do with your server resources.
Regards,
Bryle
Bryle,
the only thing I can customize is the memory limit. Unfortunately, increasing the memory limit did not help. I’m currently talking to the hoster’s support about things they might have changed which could cause the issue etc.
Thanks,
Tim
Plugin Contributor
bcrodua
(@bcrodua)
Hi Tim,
Just checking back if you have already resolved the issue with your host?
Please let me know.
Thanks,
Bryle
Hi Bryle,
since I use UpTimeRobot to generate some traffic during nighttime, the backup sometimes completes successfully after > 10 hours (time needed before the problems started: a couple of minutes). From what I can see from the logs, most of the time the process waits to be rescheduled. Looks like the actual problem is the upload itself which needs to be rescheduled multiple times per file.
Thanks,
Tim
One additional remark: If it does not complete, but stops after a few hours, it seems to be caused by an overlapping resumption (see log below). Anyway, the root cause seems to be that the actual upload to Dropbox is remarkably slow and somehow broken.
23623.971 (57) Requesting backup semaphore lock (86cf888d6364)
23623.972 (57) [Lock option (updraft_lock_86cf888d6364, wp_options) already existed in the database] debug
23623.973 (57) [Lock (updraft_lock_86cf888d6364, wp_options) could not be acquired (it is locked)] info
23623.973 (57) Failed to gain semaphore lock (86cf888d6364) - another resumption for this job is apparently already active
23623.974 (57) Failed to get backup job lock; possible overlapping resumptions - will abort this instance
Plugin Contributor
bcrodua
(@bcrodua)
Hi Tim,
Apologies for the the delay.
Please can you provide me full copy of the backup log? So I can check further.
Thanks,
Bryle
Hi Bryle,
sure – https://pastebin.com/RBKbJdFs
Thanks,
Tim
Hi Bryle,
it pretty much looks like as if the problem disappeared for no reason (no relevant change on my side). One scheduled and one manual backup completed successfully and very fast. I guess the problem has been caused by some sort of (networking?) issue at the host which has been resolved in between.
Thanks a lot for your support!
Tim
Hi Tim,
Semaphore locks depend upon accurate time, and break if the time goes significantly backwards; so it’s possible that your webserver had inaccurate time, and the server admin then spotted and fixed it.
David
