[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: ext3 performance w.r.t. journal size

Ted writes:
> The mke2fs/tune2fs programs have an upper limit of 100MB (and a lower
> limit of 1MB --- but I suspect 1MB journal using data journaling won't
> be much fun :-).

Actually, the lower limit is 1024 blocks.  Even so, in my testing with
a 1kB block ext3 filesystem + 1MB full-data journal, it was only about
15% slower than a 4kB block filesystem + 32MB full-data journal on bonnie++
block I/O speed.  In comparison, a 4kB block with a 4MB journal was
about 95% of the 4kB block + 32MB journal filesystem.

I'm not sure how much of the 1MB journal vs 4MB journal speed difference
was simply due to the fact that a 1kB block filesystem is slower than a
4kB block filesystem.  I can't test it right now since I've got some
debugging on in my kernel, but I have a whole series of benchmarks on
the same system, so I will get some numbers for pure 1kB vs 4kB block
performance on ext2 and ext3 as well.

Cheers, Andreas
Andreas Dilger  \ "If a man ate a pound of pasta and a pound of antipasto,
                 \  would they cancel out, leaving him still hungry?"
http://www-mddsp.enel.ucalgary.ca/People/adilger/               -- Dogbert

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]