Hello, everyone.
On the eos system we are running, we have noticed that about once or twice a month, one FS will get an opserror and automatically drain.
Most of the time when we get this error, we don’t find anything wrong when we look at the disk.
I was wondering if the way to detect opserror is to write “scrub.rewriteXX” files to the FST disk volume and detect it when it fails to write.
I was wondering how many errors it takes to detect it as an opserror and if we can change the setting if we think it is too sensitive.
Regards,
– Geonmo