Upgrade 26>27 : Yet another "Step 6 is currently in process. Please reload this page later." #547
Comments
Keep in mind you can always do a manual update. There's no need to stay down while troubleshooting a situation like this: https://docs.nextcloud.com/server/latest/admin_manual/maintenance/manual_upgrade.html
Does the parent folder ( That step does three things:
Can you elaborate on the types of issues?
That PR just expanded the error message to make it easier for people to reset things if they wanted to re-trigger the Update run. This is just a generic log entry. There's no universal "old well known issue" here since a million things can cause any given step to fail. Have you tried running the Updater rom the command line? In addition to being more reliable, there's a chance there some error output not being shown in the log. Here's how to run the Updater in command line mode:
Can you share a bit more about the underlying system? Hardware/VM? Filesystems in use? |
|
Hello Josh, Thank you for your help.
NO. There's only one folder in it. so you seem to confirm that he zip file was properly extracted and deleted. I have tried to run the updater from the command line but I got the same message. I have not tried yet after having deleted the .step file The thing is: since my nextcloud system is actually in an unstable state, I don't want to make things worst and unrecoverable by launching commands for which I don't fully understand the impact. I wont elaborate on this because I just cant remember any detail but I am pretty sure that I did triggered more issues than solutions when trying occ command on a failed upgrade some years ago (I am using Owncloud then Nextcloud at fork for many years) It is a vmware VM from Esxi 6 running on a Dell server (and the cloud is only a part of everything it runs (fine) The nextcloud is on /dev/sdb1 If that info helps : I was previously (so yesterday) on v25 and did upgrade to v26. It worked. Then I tried right away to upgrade to v27 (and everything from ACP in the web interface) |
Please do so if possible. The Updater from the command line is the same underlying Updater as web one. The worst that should happen - and what I expect will happen in your case - is that it will get stuck at the same step. What I'm hoping for though is that there will be some clue output to stderr that you're not seeing when you run it via the Web.
I understand the caution. I'm not suggesting any Something very strange is going on here because the code that is in that step right after removing the ZIP file is super simple. It loads two tiny text files (the For completeness, maybe share the content of your |
|
@joshtrichards then we might have spotted the issue but not the root cause yet. I dont have any file /var/www/html/nextcloud/version.php !!! This is the contain of my nextcloud root dir :
I thus dont touch anything and wait for your next guidelines |
|
just in case : I am absolutely certain that I haven't done anything manually that could have deleted files in the root folder. It should have happened during an upgrade |
!!! The screenshot looks like how it might if you somehow managed to get to later steps. I suspect if you run at the command line you'd get an exception telling you that the version file is missing. We should probably catch that better to report it in web mode, but that still leaves is with the mystery of how you ended up in this state. The step that would do this would be
|
So you can see that its there in the very first nextcloud-26.0.13.1-xxxxxxx folder but not the last. I must say that I am unsure about the exact symptoms on that attempt because you can see that I retried hard afterwards. But I dont remember having had something different than finally stuck on 6th step. It should be noted also that on all attempts the script was also timing out on DB backup step and I had to refresh or come back later, which finally worked. And I do have a pretty powerful and fully dedicated MySQL server running on another Ubuntu VM and only on SSD. So a DB dump should normally be fast on my system (I only have a few users and a few thousands of files) Updater log attached |
|
Thanks - that is helpful. Though there is still a mystery here. Things went wrong at some point during the first run attempt to update from 26.0.13.1 -> 27.1.8.1, but the source is unclear. This was the run with the good backup: Everything look good up until That indicates that your installation folder was wiped out somehow between when the backup step ran (because we know the files were there at that point - they're in your backup!) and the above step. But we haven't deleted anything yet at this point so I have zero idea how that would have happened. I understand why things stopped where they did, but not how you ended up in this state. The above error is just a symptom of whatever the real culprit is: the files for your active 26.0.13.1 installation were removed/disappeared from the installation folder at some point after we did our backup. There are no indications that anything unusual happened in between per the logs or file listings you've provided. If I had to take a wild guess, maybe something happened during this big time gap here: Not that there's necessarily anything wrong with a huge time gap, but it is... suspicious. Did the server go down or was it rebooted or any restores from backups or manual interventions you made during that time that you can think of? Here's the full excerpt from your log for just the specific Updater run in question: |
No. Nothing of that. This from Prometheus/Grafana running on another server and monitoring all my infrastructure. The Nextcloud disk is the sdb (yellow/blue). And we can see some R/W activity on it around 12:00, and then around 17:00, which is excatly what you have found in the logs. Between this timeframe I just let the upgrade run because it was timed out, and spend my afternoon on something else. Which is according to me, the wiser thing I could have done. Its only after that I tried hard to reinstall the upgrade. Just FYI the violet is the /media disk and my seedbox server was probably moving a lot of downloaded files on it until 14:30. Just everything normal on this specific server that is beefed for streaming media to friends and family. And this clearly helps to prove that there were no downtime or issue at all on it. Now the question is : what do you propose to do so I can finally have a cloud again ? :-) |
I'd suggest the following, ranked by what I'd probably personally do:
Choice 1 is easy enough to try. If it fails, just jump to choice 3 unless you feel like experimenting (because it's possible choice 2 will turn up some new info, but I'm also somewhat doubtful). |
joshtrichards
added
feature: deletion
|
I chose the 1. The problem is that when I launched the upgrader it was still bugging on : Which is logical as the .step file was still on "start step 6" I backuped the .step file and try a first attempt changing it to "end step 5" So what I did is just changing the .step file again to "end step 6"... and there it worked After that I disabled the maintenance mode manually in the config and went to the upgrader web page which said complete, and I ran the rest of the upgrade process in the GUI and everything worked All in one all this convinced to me to finally move my nextcloud install to full docker to avoid those upgrade issues in the future. Not done yet but working on it Thank you for your assistance ! |
Stuck forever at "Step 6 is currently in process. Please reload this page later."
I let it on the step for several hours and it does nothing, pages refresh always show this
This is an old well known issue that was supposedly fixed there, but it still happen on my 26.0.13.1 installation. I read all the reports and tried mostly everything and nothing helped.
I have deleted multiple times the /data/update/.step file to redo everything again and it's always stuck there.
I am on Ubuntu 220.4 with Apache and PHP up -to-date (8.2)
the PHP memory limit is set to 512MB
Updater log file shows nothing helpful
Contain of the downloads/nextcloud folder shows that it did extract the zip so why doesn't it just proceed with the next step of the installation ?
My cloud is down for 24 hours and I just don't know what to do to fix it.
since I had all sorts of worst issues in the past with the occ maintenance_fix thing, I am reluctant to try to solve my issue with it and I would like to have some guidance on what to do.
The text was updated successfully, but these errors were encountered: