ceph charms should use the CFQ IO scheduler so scrubbing can be de-prioritized
Bug #1624591 reported by
James Troup
This bug affects 2 people
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Ceph OSD Charm |
Triaged
|
Medium
|
Unassigned | ||
OpenStack Ceph Charm (Retired) |
Won't Fix
|
Medium
|
Unassigned | ||
ceph (Juju Charms Collection) |
Invalid
|
Medium
|
Unassigned | ||
ceph (Ubuntu) |
Invalid
|
Undecided
|
Unassigned | ||
ceph-osd (Juju Charms Collection) |
Invalid
|
Medium
|
Unassigned |
Bug Description
Since firefly Ceph can set the IO priority and class for scrub traffic; this would help immensely on busy Ceph clusters. It does require using the CFQ scheduler for OSDs but this should be fine since OSDs are (generally) the only user of that device.
More information here: http://
Changed in ceph-osd (Juju Charms Collection): | |
status: | New → Triaged |
Changed in ceph (Juju Charms Collection): | |
status: | New → Triaged |
importance: | Undecided → Medium |
Changed in ceph-osd (Juju Charms Collection): | |
importance: | Undecided → Medium |
Changed in charm-ceph: | |
assignee: | nobody → Chris MacNaughton (chris.macnaughton) |
importance: | Undecided → Medium |
status: | New → Triaged |
Changed in ceph (Juju Charms Collection): | |
status: | Triaged → Invalid |
Changed in charm-ceph-osd: | |
assignee: | nobody → Chris MacNaughton (chris.macnaughton) |
importance: | Undecided → Medium |
status: | New → Triaged |
Changed in ceph-osd (Juju Charms Collection): | |
status: | Triaged → Invalid |
Changed in charm-ceph: | |
assignee: | Chris MacNaughton (chris.macnaughton) → Chris Holcombe (xfactor973) |
Changed in charm-ceph-osd: | |
assignee: | Chris MacNaughton (chris.macnaughton) → Chris Holcombe (xfactor973) |
Changed in ceph (Juju Charms Collection): | |
assignee: | Chris MacNaughton (chris.macnaughton) → Chris Holcombe (xfactor973) |
Changed in ceph-osd (Juju Charms Collection): | |
assignee: | Chris MacNaughton (chris.macnaughton) → Chris Holcombe (xfactor973) |
Changed in charm-ceph: | |
assignee: | Chris Holcombe (xfactor973) → nobody |
Changed in charm-ceph-osd: | |
assignee: | Chris Holcombe (xfactor973) → nobody |
Changed in ceph (Juju Charms Collection): | |
assignee: | Chris Holcombe (xfactor973) → nobody |
Changed in ceph-osd (Juju Charms Collection): | |
assignee: | Chris Holcombe (xfactor973) → nobody |
tags: | added: canonical-bootstack |
To post a comment you must log in.
This could be done on all machines that house disks by setting the system scheduler to cfq, it looks like changing it per-disk, permanently will be a bit of a hack with a service to write `cfq` to the scheduler file. On the other hand, according to this (http:// www.ilsistemist a.net/index. php/linux- a-unix/ 38-linux- i-o-schedulers- benchmarked- anticipatory- vs-cfq- vs-deadline- vs-noop. html) somewhat older benchmark, cfq vs deadline isn't a huge performance killer unless using direct-io in which case deadline performs a bit better.
@elmo how would you feel about the ceph[-osd] units having the default scheduler be cfq instead of deadline?