Automate recovery of App/services after an OOD condition
-
Out of Disk (OOD) condition is when the OS has no more room to write any files, and any files it tries to write to with existing data end up 0 byte files.
This traditionally means manual cleanup of disk space, which may still be required, but furthermore hunting for any applications and services that fail to start because their config files have been corrupted or overwritten as zero byte files.
Automating this process gets Cloudron and its services back up faster and more reliably.
-
Are you referring to the recent post(s) about the certificate issues?
There was a bug that nginx configs of stopped apps was not cleaned up, this was fixed a couple of releases ago (but some of the Cloudrons did not update hitting the issue).
A separate bug is when things run out of disk space. it sometimes ends up in 0 byte files. But this is not limited to just cert files but all over. That requires a deeper fix.
-
Are you referring to the recent post(s) about the certificate issues?
There was a bug that nginx configs of stopped apps was not cleaned up, this was fixed a couple of releases ago (but some of the Cloudrons did not update hitting the issue).
A separate bug is when things run out of disk space. it sometimes ends up in 0 byte files. But this is not limited to just cert files but all over. That requires a deeper fix.
