ADSM-L

Re: Optimum FS size, performance breakpoints?

1999-05-05 15:19:09
Subject: Re: Optimum FS size, performance breakpoints?
From: "Allen S. Rout" <asr AT NERSP.NERDC.UFL DOT EDU>
Date: Wed, 5 May 1999 15:19:09 -0400
=> On Wed, 28 Apr 1999 09:30:13 -0400, Richard Sims <rbs AT BU DOT EDU> said:

>> We're about to move the data on our campus mail server (45000 users, ca 60G
>> of mail) to a new hunk of disk, and are evaluating our filesystem
>> arrangement.

> You didn't say what your mail server software is or file system topology,
> but some general feedback...
>
> [ excellent basic tuning data ]
>

Thanks for that note.  We're trying to tune at a somewhat finer granularity.

We use the cyrus IMAPD server, which stores mail in a structure grossly
similar to the Rand MH system. The details differ, but it maintains a "folder"
structure with heirarchical directories. Each message uses a separate file;
this is dramatically different from Sendmail's storage format.

Number of files per filesystem is really the value I'm trying to tune now.
Since we have so many, it really appears that this value, and its impact on
database transfers (for normal incrementals) and updates (as incrementals run)
is becoming a significant issue.

Since the number of files in a _directory_ is going to be more-or-less
completely under the control of the users (i.e. how often they clean stuff
up), our only real tuning knob is how many users we place on a filesystem.

I was hoping someone had noticed something on the order of "We started seeing
slowdown when we exceeded 200K files on a FS", or some such.


Allen S. Rout
<Prev in Thread] Current Thread [Next in Thread>