[ale] kjournald?

David Tomaschik david at systemoverlord.com
Mon Sep 19 18:58:24 EDT 2011


On 09/19/2011 06:39 PM, Jim Kinney wrote:
>
> Is it one journal write per file or per inode used per file?
>
> On Sep 19, 2011 4:24 PM, "David Tomaschik" <david at systemoverlord.com
> <mailto:david at systemoverlord.com>> wrote:
> > On Mon, Sep 19, 2011 at 3:19 PM, John Heim <john at johnheim.net
> <mailto:john at johnheim.net>> wrote:
> >>        What (if anything) does it mean if iotop shows that the process
> >> doing the most disk I/O is kjournald? Actually, on 4 partitions
> that I have
> >> on an ISCSI array, the kjournald process is the top I/O on each.
> >>
> >> I'm thinking of setting the noatime option in my fstab.
> >
> > It means that the journal on those disks is being written to a lot.
> > Typically, you'll see this with lots of small writes. Each write
> > requires the following:
> >
> > 1) Write change to journal.
> > 2) Write change to filesystem.
> > 3) Mark change as done in journal.
> >
> > So that's 2 journal writes for every FS write.
I'm not sure, to be honest.  But I suspect it doesn't matter, as the
multiple journal entries are likely to fit in a single block and form
one logical journal write anyway.

John -- if you don't need atime, then noatime does offer a performance
benefit.  Alternatively, relatime is a nice middle-ground:

            Update inode access times relative to modify or change
time.  Access time is only  updated  if  the  previous
              access  time  was earlier than the current modify or
change time. (Similar to noatime, but doesn't break mutt
              or other applications that need to know if a file has been
read since the last time it was modified.)

              Since Linux 2.6.30, the kernel defaults to the behavior
provided by this option (unless noatime  was   speci-
              fied),  and  the  strictatime  option  is  required to
obtain traditional semantics. In addition, since Linux
              2.6.30, the file's last access time is always  updated 
if  it  is more than 1 day old.

(Assuming, of course, your kernel is older than 2.6.30.  Otherwise, as
stated above, you already have relatime.)

-- 
David Tomaschik, RHCE, LPIC-1
System Administrator/Open Source Advocate
OpenPGP: 0x5DEA789B
http://systemoverlord.com
david at systemoverlord.com

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.ale.org/pipermail/ale/attachments/20110919/0e237e99/attachment.html 


More information about the Ale mailing list