Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 4 Feb 2003 00:07:52 +0300 (MSK)
From:      Andrey Alekseyev <uitm@zenon.net>
To:        Ryan Dooley <ryan@third-man.com>
Cc:        Bill Moran <wmoran@potentialtech.com>, stable@freebsd.org
Subject:   Re: recommendations on the newfs of a 1.0TB fs...
Message-ID:  <200302032107.h13L7qh82700@uitm.zenon.net>
In-Reply-To: <20030203204524.GB56152@elvis.mu.org> from Ryan Dooley at "Feb 3, 2003 12:45:24 pm"

next in thread | previous in thread | raw e-mail | index | archive | help
> I'm wondering what values I might try for -i that might be reasonable.

I believe, you should examine your existing file system more closely.

Recently we had a task of optimizing a large filesystem (300+ GB) for our
large mail cluster (hundreds of thousands of mail accounts and about 10000+ mail
domains). The filesystem itself was supposed to contain about 1500000
small files and about 400000 directories.

For that task I had written a rather simple utility that traverses a
directory tree and produces a statistics report on the number of files for
each given range of size (less than 8KB, 8KB < n < 16KB, greater than
16KB, etc.). Also, the total number of files and directories is accumulated
along with other interesting detail. I must say it was rather a useful
information and food for thought before doing any optimization! :)

However, I should mention, the OS is Solaris 8 and the filesystem is not UFS. 
I was optimizing given certain details on file system allocation mechanisms.
Especially, block size was the question.


-- 
Andrey Alekseyev. Zenon N.S.P.
Senior Unix system administrator

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-stable" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200302032107.h13L7qh82700>