Type | Version/Name
--- | ---
Distribution Name | any
Distribution Version | any
Linux Kernel | any
Architecture | any
ZFS Version | any
SPL Version | any
currently I have arcstat, arc_summary, zpool status and zpool iostat (please correct me if there are more tools)
It would be awesome to combine all information (including top-like reports for zfs kernel modules and processes) into a single dashboard like atop or nmon does for the os in general. ideally tunables should be shown in a separate view like powertop does.
there is also dstat --zfs-arc --zfs-l2arc --zfs-zil
Such a thing would be nice, especially if it could display IO load and latencies on a _per-dataset level_.
Would be nice, if we can view which VM on which ZVol is causing issuing how much I/O to ZFS in which size etc. and how many real disk reads/writes result from this, also the Arc hitrate would be interesting per ZVol as well as the CPU-Usage from ZFS per ZVol.
A backlog-count in the log-device per ZVol would also help to debug excessive loads on these devices. :)
thanks, amazing to see this feature has been solved!
Most helpful comment
there is also
dstat --zfs-arc --zfs-l2arc --zfs-zil