Control: tags -1 unreproducible On Wed, 2021-05-05 at 17:27:58 +0200, Juergen Richtsfeld wrote: > Package: prometheus-node-exporter > Version: 1.1.2+ds-1+b1 > Severity: normal > X-Debbugs-Cc: juergen.richtsf...@gmx.at
> I tried to get disk (filesystem) i/o metrics like > node_disk_read_time_seconds_total from the diskstats collector but it > doesn't seem to provide any (there is not a single metric having a name > starting with node_disk). I don't find errors in the log (there is none) > and in the journal. I already tried to get rid of the default > diskstats.ignored-devices but this didn't help. > -- Configuration Files: > /etc/default/prometheus-node-exporter changed: > ARGS="--collector.diskstats.ignored-devices=\"\" > --collector.filesystem.ignored-mount-points=\"^/(dev|proc|run|sys|mnt|var/lib/docker/.+)($|/)\"" On Fri, 2021-05-07 at 19:05:56 +0200, Jürgen Richtsfeld wrote: > I just found that if I put "--collector.diskstats" in the ARGS of > /etc/default/prometheus-node-exporter and restart the service it works. > I didn't try this i in the first row as the documentation says that this > is default enabled anyways (so the bug basically just is, that the > description about the default is wrong!). Well, I cannot reproduce, the collector is enabled by default and it works out of the box. Package installed freshly on a system: # grep ARGS /etc/default/prometheus-node-exporter ARGS="" # GET http://localhost:9100/metrics | grep ^node_disk_ | wc -l 68 Given this, I'm assuming there's perhaps some local misconfiguration going on, or some other thing preventing this to work. Thanks, Guillem
