No disk tabs shown in disk usage nor disk I/O

  • Neither the "Disk usage" nor the "Disk I/O" pages list any of my drives and are therefore completely empty:



    This problem seems to have started after I recently upgraded from OMV4 to OMV5 in July 2020, using the instructions from this post.


    Graphs are generated under /var/lib/rrdcached/db/localhost/disk-sd? folders. For example:

    Code
    root@Sirius:~# ls -alh /var/lib/rrdcached/db/localhost/disk-sda
    total 1.6M
    drwxr-xr-x  2 root root 4.0K Mar 31  2018 .
    drwxr-xr-x 27 root root 4.0K Jul 10 14:02 ..
    -rw-r--r--  1 root root 289K Jul 25 18:04 disk_io_time.rrd
    -rw-r--r--  1 root root 289K Jul 25 18:04 disk_merged.rrd
    -rw-r--r--  1 root root 289K Jul 25 18:15 disk_octets.rrd
    -rw-r--r--  1 root root 289K Jul 25 18:15 disk_ops.rrd
    -rw-r--r--  1 root root 289K Jul 25 18:15 disk_time.rrd
    -rw-r--r--  1 root root 146K Jul 10 13:58 pending_operations.rrd

    I use openmediavault-snapraid and openmediavault-unionfilesystems plugins.

    OMV v5.5.4-1

    Kernel 5.6.0-0.bpo.2-amd64


    I did find an issue describing this same problem here and added a comment, but that issue was closed suggesting the drives should appear after a while. This is not the case: they don't show up, even after waiting for several hours.


    Ideas?

    NAS: Celeron G1840T | ASUS H87I-Plus | Samsung SSD 850Pro 256 GB | 5xWD40EFRX | OMV6

  • Not much luck getting this to work yet...


    Note that ZFS is not used here (others have reported issues in the past with the same graphs while using ZFS).

    NAS: Celeron G1840T | ASUS H87I-Plus | Samsung SSD 850Pro 256 GB | 5xWD40EFRX | OMV6

  • Found that the collectd "disk" plugin was not executable:

    A chmod 755 /usr/share/openmediavault/mkrrdgraph/plugins.d/disk.py fixed that, but the disks still don't appear in OMV (even after running omv-mkrrdgraph and rebooting)...


    Still looking.

    NAS: Celeron G1840T | ASUS H87I-Plus | Samsung SSD 850Pro 256 GB | 5xWD40EFRX | OMV6

  • In the meantime, I can confirm the graphs themselves are actually generated: https://myip/rrd.php?name=disk-time-sda-day.png does show an (up-to-date) graph. That would suggest the problem is with the enumeration of the disks (as you seem to think as well?!).

    NAS: Celeron G1840T | ASUS H87I-Plus | Samsung SSD 850Pro 256 GB | 5xWD40EFRX | OMV6

    • Offizieller Beitrag

    The problem is that your RPC response is a dict/object, but it must be an array. I have really no idea how this can happen.

  • Found the cause: I had an entry as follows in config.xml

    Code
    <mntent>
      <uuid>e8ab8481-4f9a-4ecb-a439-9fc5a12ea10f</uuid>
      <fsname>/srv/86531979-48b9-42be-a7d1-e0d735bd7880</fsname>
      <dir>/srv/86531979-48b9-42be-a7d1-e0d735bd7880</dir>
      <type>fuse.mergerfs</type>
      <opts></opts>
      <freq>0</freq>
      <passno>0</passno>
      <hidden>1</hidden>
    </mntent>

    Note the bad fsname value. Problem went away when I replaced it by 86531979-48b9-42be-a7d1-e0d735bd7880:

    I'm guessing I mis-configured this when I originally set up mergerfs on my system (I vaguely remember having edited config.xml at the time). It appears this messed up the enumerateMountedFilesystems RPC.


    Thanks for the help!

    NAS: Celeron G1840T | ASUS H87I-Plus | Samsung SSD 850Pro 256 GB | 5xWD40EFRX | OMV6

  • apa

    Hat das Label gelöst hinzugefügt.

Jetzt mitmachen!

Sie haben noch kein Benutzerkonto auf unserer Seite? Registrieren Sie sich kostenlos und nehmen Sie an unserer Community teil!