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

Re: using non-default blocks per group

On Fri, Oct 19, 2001 at 03:17:46PM -0700, Jeffrey W. Baker wrote:
> I have always used the -R stride option but it never works.  I always get
> a metadata hotspot.  For example, copying the mozilla source tree
> recursively will result in twice as many SCSI commands on the hot disk as
> the other disk (in a 2-disk RAID 0).
> I posted over on the linux-raid list about this.  I tried a chunk of 128k
> and a stride of 8, 16, 32, and 64 with a 4k block and all configurations
> had a hot spot.  By changing the blocks per group, traffic was balanced
> evenly.

I'd have to look at the dumpe2fs output and (ideally) an trace of the
disk activity to be certain, but I'm pretty sure that what's going on
is that the hot spot is caused by the inode table.  The -R stride
option will distribute the block and inode bitmaps across the various
disks.  But with a chunk size of 128k, if the average block group has
less than 1024 inodes, all of the inodes for that block group will be
in one chunk, and if the chunk size evenly divides the block group
size, then on a two disk RAID system, all of the inodes accesses will
be concentrated on one disk.

So adjusting the block group size may be the best bet for dealing with
your problem.

							- Ted

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