Hi bonienl, the s3 plugin no longer shuts down my system ever since updating to the latest version. In the syslog I see nothing but "Disk activity ongoing" entries, every time for the first disk in my array, namely the parity disk.
The changelog indicated "disk state" is now considered in the disk activity check. Does this mean the disks have to be spun down in order to meet the requirements? From my observations that's exactly what happens.
If this is intended, then this is a breaking change for me and anyone else who does not allow one or more of the disks considered to spin down. It also makes little sense for it to still check disk I/O as well if that's the case. A better solution would be to fall back on disk I/O if a particular disk is set to never spin down or add an option so we can change this behavior.
But if this is unintended, then I'm guessing this block is responsible:
diskio=($(grep -Pho "^$dev \K\d+" /dev/shm/1 /dev/shm/2))
if [[ -n $active || ${diskio[0]} != ${diskio[1]} ]]; then
result=1
break;
fi
You probably meant to do something like this:
if [[ -z $active ]]; then
continue;
fi
diskio=($(grep -Pho "^$dev \K\d+" /dev/shm/1 /dev/shm/2))
if [[ ${diskio[0]} != ${diskio[1]} ]]; then
result=1
break;
fi
That would fix the issue.