Open echan00 opened 4 years ago
So it's still working correctly sometimes? Are you running multiple instances of libreoffice at the same time?
Yes it's working most of the time.
Yes I believe I'm running multiple instances of libreoffice through additional background workers
Limiting background workers to just one should solve this. Last time I checked this seemed to be an issue with libreoffice itself.
But doesn't this limit the scalability of my server? I spin up new background workers when there are more documents in my queue requiring libreoffice file conversion..
Sure it does. Getting past this requires somebody to figure out how to run parallel headless libreoffice instances successfully. The issue should persist on local environments too, so feel free to investigate.
Thanks for the tip. I'll look into this and revert.
I imagine background workers are parallel instances of their own already. Don't they each have their own Java runtime?
Starting today I have been seeing this issue (seemingly 100% of the time, not intermittent), for some reason. Still looking into it.
Starting today I have been seeing this issue (seemingly 100% of the time, not intermittent), for some reason. Still looking into it.
Well, now I'm not sure if the Java thing was just a red herring. My message was:
javaldx: Could not find a Java Runtime Environment!\nWarning: failed to read path from javaldx\nError: source file could not be loaded
But, I think that actualy always shows up to stdout (from trying soffice
in heroku run ... bash
), the important part of my message was: Error: source file could not be loaded
, which is unrelated to Java. So the problem was something else, in my case.
@Aesthetikx @echan00 any solution for this?
I am receiving the error above every so often when I use Libreoffice to convert DOCX to HTML. Anybody know what might be causing this?