Re: Some very thought-provoking ideas about OS architecture.

Chris Adams (cadams@ro.com)
21 Jun 1999 10:31:21 -0500

Once upon a time, Malcolm Beattie <mbeattie@sable.ox.ac.uk> said:
>Please don't guess/exaggerate: I've already posted fsck benchmark
>times to this mailing list a month ago. In particular, fsck for a
>43 GB ext2 filesystem (4K blocks) with 30 GB in use took 13 minutes.
>
>Disk usage was 25 directories each with a copy of the Linux 2.2.1
>source tree (63MB, 4000 files) and each with 200 subdirectories
>holding 5 files of 1MB.

That is a nice benchmark, but like all benchmarks, it is artificial and
not necessarily representative of the real world.

My news spool is a 30G RAID 0 striped across 7 ultra wide SCSI Seagate
Barracudas on a DPT SmartRAID IV hardware RAID controller (which has
performance problems with RAID 5 according to some, but not with RAID
0). It is one big ext2 filesystem with 4K blocks. Last time I had to
fsck it, it had about 22G of data in about 3,000,000 files (I had no
idea how many directories, probably around 55,000). It is on a Pentium
Pro 200 with 320MB RAM. The fsck took about 50 minutes and there were
no errors. There have been some times when the system crashed and we
had to give up on fsck and just mke2fs the news spool.

There are other ways of storing news instead of one article per file,
and they work fine for news feeding, but they still have some problems
with news reading, so I don't use them. I'm not claiming that news is
representative of all large filesystems, but I am claiming that your
benchmark is not representative either.

-- 
Chris Adams <cadams@ro.com> - System Administrator
Renaissance Internet Services - IBS Interactive, Inc.
Home: http://ro.com/~cadams - Public key: http://ro.com/~cadams/pubkey.txt
I don't speak for anybody but myself - that's enough trouble.


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/