[262] in arla-drinkers
Re: Chunk size?
daemon@ATHENA.MIT.EDU (Love)
Wed Sep 16 04:53:25 1998
From owner-arla-drinkers@stacken.kth.se Wed Sep 16 08:53:24 1998
Return-Path: <owner-arla-drinkers@stacken.kth.se>
Delivered-To: arla-drinkers-mtg@bloom-picayune.mit.edu
Received: (qmail 11347 invoked from network); 16 Sep 1998 08:53:23 -0000
Received: from unknown (HELO sundance.stacken.kth.se) (130.237.234.41)
by bloom-picayune.mit.edu with SMTP; 16 Sep 1998 08:53:23 -0000
Received: (from majordom@localhost)
by sundance.stacken.kth.se (8.8.8/8.8.8) id KAA24602
for arla-drinkers-list; Wed, 16 Sep 1998 10:47:07 +0200 (MET DST)
Received: from elixir.e.kth.se (elixir.e.kth.se [130.237.48.5])
by sundance.stacken.kth.se (8.8.8/8.8.8) with ESMTP id KAA24598
for <arla-drinkers@stacken.kth.se>; Wed, 16 Sep 1998 10:47:02 +0200 (MET DST)
Received: from hummel.e.kth.se (hummel.e.kth.se [130.237.43.135])
by elixir.e.kth.se (8.8.7/8.8.7) with ESMTP id KAA16248;
Wed, 16 Sep 1998 10:47:00 +0200 (MET DST)
Received: (lha@localhost) by hummel.e.kth.se (8.8.7/8.6.6) id KAA25268; Wed, 16 Sep 1998 10:46:50 +0200 (MET DST)
To: Dr A V Le Blanc <LeBlanc@mcc.ac.uk>
Cc: arla-drinkers@stacken.kth.se
Subject: Re: Chunk size?
References: <19980916071304.C13657@afs.mcc.ac.uk>
From: Love <lha@elixir.e.kth.se>
Date: 16 Sep 1998 10:46:50 +0200
In-Reply-To: Dr A V Le Blanc's message of Wed, 16 Sep 1998 07:13:04 +0100
Message-ID: <amd88wla11.fsf@hummel.e.kth.se>
Lines: 28
X-Mailer: Gnus v5.5/Emacs 20.2
Sender: owner-arla-drinkers@stacken.kth.se
Precedence: bulk
Dr A V Le Blanc <LeBlanc@mcc.ac.uk> writes:
> I gather from what documentation there is, that arla does not have
> a chunk size as AFS does, but that entire files are transferred
> and cached. Is this true,
> or are large files broken up into
> chunks?
Whole files. (The simple way)
> If so (or if not) is there a limit to the amount that
> can be transferred at once? Can this be configured?
Your cache size limits the files you can edit.
> Secondly, there is with Transarc AFS the possibility of having
> the cache entirely memory resident (if you have enough memory,
> and if that's what you want). But arla does not seem to be
> able to do this. Would this be a difficult alteration?
The simple way (on *BSD) is to have the cache in mfs.
Xfs doesn't really do any read/write operation by it self, so you would
have to implemement those. Then the memory need to be writeable (in some
way) for arla to store the files in. Guess it could get messy. I dont
see any quick way of doing this right now.
Love
