My backups used to work fine, and I even had scheduled backup enabled. But now, my backup zips don’t complete on my live site. They get stuck at what appears to be the end of the zip backup process - at about 320mb, which is roughly what the completed zips are, but now, the zip never completes so you get the process random text partfile name. If I edit the zip filename and remove the random part file name, they stop the process and appear completed. But I can’t open them, because the zip is empty (see first image).
The issue you’re experiencing might be related to server limits. A good starting point is to check the following limits in your server’s PHP configuration.
thanks for reply. Which limits should I look for? Is this in php info?
An error message flashes up in an interface ‘popup’ pane, with “error 500 failed to fetch …” but it doesnt stay visible very long and not all of the message is visible so I cannot see what the error actually is.
its this, according to the popup, which I just managed to copy
Fetch Failed:
Error 500 - Internal server error
Error 500 - Internal server error
An internal server error has occurred!
Please try again later.
I’ll see if I can access server logs to get more info.
ALSO may be relevant:
I deleted and replaced the admin plugin bc there’s a critical error and other warnings reported in my grav logs that say admin plugin not found, when it is there and enabled and I can use the dashboard normally.
I have cleared the caches via the admin dashboard several times.
November 18th at 4:49pm WARNING Plugin 'admin' enabled but not found! Try clearing cache with `bin/grav clearcache`
November 18th at 4:49pm WARNING Plugin 'admin' enabled but not found! Try clearing cache with `bin/grav clearcache`
November 18th at 4:49pm WARNING Plugin 'admin' enabled but not found! Try clearing cache with `bin/grav clearcache`
November 18th at 4:49pm CRITICAL Class "Grav\Plugin\Admin\Admin" not found
As @Karmalakas mentioned, check the server logs. If you have a hosting account, you’ll most likely find access to the logs in your account panel. You can also reach out to your hosting provider’s support to ask for access to server logs or details about limits on your account. A 500 error is a general error message, and there can be many possible causes for it.
thanks for reply. I checked my server logs (in Ionos these are named access logs). I looked at yesterdays and current file and searched in notepad for ‘500’, ‘Internal’ and ‘Error’ and found nothing.
In case its relevant, I looked at grav config/ php info local & master settings and max execution time is set to 50000, and memory limit is ‘-1’. That last seems weird. My master php info memory limit is set at 128mb. Just checked, ‘-1’ means set to no limit.
In my humble opinion, server errors are not logged in the access logs, which is why you can’t see them there. I am not familiar with the solution used by Ionos. In my opinion, you should contact them for additional information (you are paying them for this).
yeah I think youre right. tbh Ionos are pretty basic but I usually only need basic LAMP and they give me a very good price. I cant see any other logs except traffic, sftp and access. This is via sftp directory not any admin panel thing. Im also digging around and seeing I have internal/timeout errors from the Page Stats plugin so this might be a problem with some other sitewide script. Thanks for trying to help.
Im trying to trace the offending folder/file.
atm I think its this that fails to load.
https://.../admin/backup.json/id:0/task:backup/admin-nonce:8631f5ec09c913b225b44e9ab22f9456
(or other similar assigned numbers)
Im doing a bit by bit re-install of grav 1.7.48 because I think its the install that might be a problem. The backup system itself works fine because I tested various backup folder configs and they work. Eg the user folder alone, or all the system, vendor, bin folders, loose files etc etc.
Updating in case of benefit to other readers. This briefly documents the process to solve the problem. NB - I didnt go to Ionos server support bc (most) server support does not help with app instance issues. You have to figure this out for yourself.
I did not get the error with core backup feature with smaller backups but couldnt tell if it was to do with memory etc bc php info seemed to indicate it was fine.
But, I ran a local php.ini with max_execution_time = 6000; max_input_time = 5000; memory_limit = 1000M; in it. I think this made zero difference
I cleared all caches numerous times. Also using putty command line using bin/grav clear-cache and bin/grav clear-cache --all
I installed Backup Manager plugin, and also did not get the problem with smaller backups - but I got the same error with the ‘default’ backup, the same as with shipped backup feature.
checked backup yaml in user/config. Nothing wrong with it
I did more searching in the forum, looking for appearances of ‘nonce’. This returned various posts but the one that caught my eye was about which browser was being used.
I remembered that my localhost version was not throwing these weird errors
I also remembered that I was using Brave, with shields up. Localhost by default does not operate shields in Brave. So I turned off shields in the remote version of the site. When shields are up the panel looks like below image (1).
Turning off shields seems to have stopped the problem somewhat.
– Backup Manager now runs default backup smoothly and does not throw any errors.
– But, the core shipped backup default process still throws the error, even though the default backup zip now completes. It looks as if it isnt completing but checking the backup folder (using sftp for example) will show its completed.
