[systemd-devel] Journalctl Unit Completion Optmization

[Zbigniew Jędrzejewski-Szmek]

On Wed, Apr 22, 2015 at 06:37:06PM +0200, Lennart Poettering wrote:
> On Tue, 03.03.15 23:02, Justin Brown (justin.brown at fandingo.org) wrote:
> 
> > Hello,
> > 
> > I've recently been using journalctl on one of my longer-lived systems, and
> > I'm running into a slight annoyance. Most of the time I'm using `journalctl
> > -u <<unit>> -e`. I use tab completion for the unit name, and the annoyance
> > is the long pause journalctl takes to complete a unit. With cold cache
> > (meaning writing 3 to /proc/sys/vm/drop_caches), this takes 23 seconds on a
> > system using Btrfs on a SATA3 SSD. With warm cache, it still takes 3-4
> > seconds. Both cases are quite a bit slower than normal tab completion
> > behavior.
> 
> Is it really that bad if the files on btrfs are defragmented properly?
> Note that btrfs is notoriously bad at handling journald's write
> pattern currently, which results in heavily fragmented files. Use the
> "filefrag" tool to check how fragmented they are, and use btrfs defrag
> to the fix this.
> 
> Before we look into other options I'd really prefer if you could check
> if this already fixes your issues.
> 
> (Note that very recent journald will set the btrfs NOCOW flag now on
> the journal files, which should work around this issue.)
Also make sure that you're running recent enough and have commit v219-99-g950c07d421.

Zbyszek