 T4ng
|
I just did another test: on the same local environment, by re-uploading the original database, then disabling some plugins before it runs the problematic custom_title update process. And it ran successfully!
So, I wonder if:
1. it's an issue to disable plugins while the updates run?
2. If after having disabled some plugins, there is a way to resume the paused process on our live website, manually, or it it should resume automatically? So that I could fix this live website?
Thanks
|
 Kor
|
Thanks for the update. Since it worked after disabling some plugins, it seems there may be a conflict with one of them.
You can temporarily deactivate the plugins and allow the process to finish. It only needs to run once and will resume automatically until it’s completed.
Before doing this, please create a full backup and then test it on your live site. Let me know how it goes once you’ve tried it.
|
 T4ng
|
OK, but can you please let me know ***how to relaunch*** the updates?
As already explained, the Running/Pause pictogram aren't buttons.
Thanks
|
 Kor
|
Thanks for your reply. Could you share a screenshot of the admin bar and the process status that you're getting now?
|
 T4ng
|
On the local env: nothing in the admin bar anymore since the updates are completedd.
On the live env: same as before since nothing changed: the custom_title process is still paused, and I can't click on any of these processes.
How do I get it to run again?
|
 Kor
|
Thanks for your reply. Could you please back up your site and then access your database and truncate the icl_background_task table?
Let me know if this helps.
|
 T4ng
|
What's the expected behavior after truncating this table?
|
 Kor
|
Thanks for getting back to me. Running that action should clear the stuck process, and if needed, WPML will automatically run it again to finish the operation.
Please make sure to create a full backup before proceeding.
|
 T4ng
|
I've been testing what you suggested:
1. Reuploaded the initial db (before any updates)
2. Truncated the icl_background_tasks table
3. Synced ACFs
Results: there's only one update remaining process in the list, which gets quickly stuck as well (remains on the "ongoing" pictogram) - but the figures won't progress.
On top of that, if we truncate the background_tasks table, and if these tasks won't come up again afterwards, I understand that the related data won't be processed. Am I wrong?
|
 T4ng
|
I did another test.
On the local environment, deactivating Elementor is enough resolve the issue: all the processes will complete without issue as soon as Elementor is deactivated.
Therefore, I'd like to try to deactivavate Elementor on the live website to solve the issue, for the time it needs to complete.
However, to do this, I need to resume the updates.
So can you tell me how I can resume the updates, which are paused right now?
Thanks
|
 Kor
|
Thanks for getting back to me. Could you please try deactivating and reactivating the ACFML plugin to see if that resets the process? If it doesn’t help, you’ll need to proceed with the database truncation as mentioned earlier.
It’s likely that the process has already completed but the status remains stuck. You can confirm this by checking the pages using ACF and attempting to translate them.
Let me know the results.
|
 T4ng
|
As explained, disabling Elementor allowed overcoming the problematic update.
To "unlock" the process, I had to:
0. Backup the database
1. In the database > icl_background_tasks > lower the amount of attempts for the stuck process, and move the status from 2 to 1. That was the way to resume the process. I wish you guys told me that when I asked.
2. Disable Elementor Pro, then Elementor
3. Clear WPML Cache (not sure this was required)
Disabling Elementor even shortly allowed the process to resume: once the 50 went to 60 and over, I could reactivate it, it won't get stuck anymore. I don't know if it explains what was happening.
After that process, I synced the ACFs, then once all of these were completed, I processed the database optimization. Which went flawlessly
I don't know if that was the right way to process, but I guess I had to figure it myself.
|
 Kor
|
Thanks for your reply. I’m glad to hear you managed to resolve it. If the issue returns, please share access to the site so we can investigate further.
|