Stll wrong i/o scheduler used in 17.04?

[GreyGeek]

Bfq?

[oshunluvr]

http://algo.ing.unimo.it/people/paolo/disk_sched/
https://www.phoronix.com/scan.php?pa...-Replacing-CFQ

[GreyGeek]

My bad! I thought you were making a joke about some "B(trfs)IO scheduler!
I can see why it would be desirable if one was using cgroups on Btrfs.

[oshunluvr]

I'm not 100% sure, but I don't think that Cgroups (control groups for disk i/o processes) are the same a Qgroups (quota groups to limit or control disk space usage).

From the respective Wiki's:
cgroups (abbreviated from control groups) is a Linux kernel feature that limits, accounts for, and isolates the resource usage (CPU, memory, disk I/O, network, etc.) of a collection of processes.
qgroups (abbreviated from quota groups) imposes an upper limit to the space a btrfs subvolume or snapshot may consume.

[jlittle]

...move to CFQ with buffered writeback throttling (WBT) + WBT_MQ (WBT multi-queu) enabled.
We originally moved to DEADLINE because of the issues with slow I/O (say to flash drives) causing applications to hang while blocked on the slow I/O being flushed out. It seems that with the recent 4.10 WBT driver...

Thank you, that sounds good, that the problem I had back in the day has been addressed.

...Faster boots. On a 8 thread Xeon CPU E3-1275 I'm seeing a reduction in usertime boots from 33.92s (Deadline) to ~24.5s (CFQ)...

Maybe that's with a spinning disc... or not with systemd. My Kubuntu boots to the login prompt (from the Grub menu) in about 2 s.

[GreyGeek]

I'm not 100% sure, but I don't think that Cgroups (control groups for disk i/o processes) are the same a Qgroups (quota groups to limit or control disk space usage).

From the respective Wiki's:
cgroups (abbreviated from control groups) is a Linux kernel feature that limits, accounts for, and isolates the resource usage (CPU, memory, disk I/O, network, etc.) of a collection of processes.
qgroups (abbreviated from quota groups) imposes an upper limit to the space a btrfs subvolume or snapshot may consume.

No doubt that you are right, but a clever admin would probably combine the effects of the two cgroups to produce some pretty stringent regulations.