Hi, That was it! After commenting out ReadOnlyDirectories=/ netdata started reporting correct disks/partitions. I guess this issue should be discussed on netdata github issue tracker.
Thanks for pointig this out. 2017-10-24 14:28 GMT+02:00 Lennart Weller <lenn...@lennartweller.de>: > I assume it has something to do with our strict ReadOnly policy applied by > systemd. > Try changing the netdata service file (/lib/systemd/system/netdata.service) > to be more lenient. > e.g. Change ReadOnlyDirectories=/ to ReadWriteDirectories=/ or remove the > lines completely. Don't forget to reload the service file after a change > > I have honestly no idea why the daemon would need write permissions on > those directories but it's worth a try as there were some odd cases before. > If that doesn't help you can remove some of the other security related > lines and check if it helps. > > October 17, 2017 3:48 PM, "Skibbi" <ski...@op.pl> wrote: > [...] > > Manually installed netdata (from binaries) displays following partitions > under > > Disks menu: > > sda - i/o stats > > / - disk space and inodes count > > /dev - disk space and inodes count > > /dev/shm - disk space and inodes count > > /run - disk space and inodes count > > /run/lock - disk space and inodes count > > > > Debian packaged netdata displays completely different partition > configuration: > > sda - i/o stats > > /tmp - disk space and inodes count > > /var - disk space and inodes count > > /var/tmp - disk space and inodes count > > > > Debian packaged netdata unfortunately fails to detect correctly disks on > my > > servers therefore monitoring them is not very reliable and I haven't > found a way > > to fix the partitions manually in netdata config. >
