How many mailboxes could one expect on a system with the following specs
Receiving messages via LMTP and only supporting pop for retrieving email.
front-end systems would be running exim, proxypop3d, apache and DNS
back-end systems would only have cyrus. There would be a third system
to provide ev
Folks I need a bit of feedback and insight.
A co-worker of mine is implementing an IMAP server here and has decided to ditch
Cyrus after he thought it was too slow and could not configure it. That being
said, I don't want to address the configuration issue. I am interested in the
speed or perfor
Michael J Barber wrote:
> Folks I need a bit of feedback and insight.
>
> A co-worker of mine is implementing an IMAP server here and has decided to ditch
> Cyrus after he thought it was too slow and could not configure it. That being
> said, I don't want to address the configuration issue. I am
This rings a bell.
His test system is a Pentium 200 with less than 10 users and just auto generated
test messages. The mail boxes were growing to 20-30MB...at least with the
Courier setup.
Any other info would be great and thanks for the note.
BTW the eventual production system is a dual Xeon 3
Michael J Barber wrote:
> This rings a bell.
>
> His test system is a Pentium 200 with less than 10 users and just auto generated
> test messages. The mail boxes were growing to 20-30MB...at least with the
> Courier setup.
>
> Any other info would be great and thanks for the note.
>
> BTW the e
Deliver 4-5k messages to a single mailbox and then delete message
10,100,1000 etc. Do that for 1000 users at the same time.. Then Cyrus
really starts to perform!
Mike Brodbelt wrote:
>
> Michael J Barber wrote:
> > This rings a bell.
> >
> > His test system is a Pentium 200 with less than 10 use
On Tue, 24 Feb 2004, Michael J Barber wrote:
> Cyrus after he thought it was too slow and could not configure it. That being
Hah, I bet he had Cyrus on a partition (or directories) operating in sync
mode :) (I assume he had a 2.x Cyrus install).
> Has anyone done similar things and noticed one
On Tue, 24 Feb 2004, Michael J Barber wrote:
> His test system is a Pentium 200 with less than 10 users and just auto generated
> test messages. The mail boxes were growing to 20-30MB...at least with the
> Courier setup.
Oh. Yeah, that is a scenario where Courier might just be faster than Cyrus.
On Wed, Feb 25, 2004 at 01:11:30AM -0300, Henrique de Moraes Holschuh wrote:
> 3. A proper filesystem (ext2 and ext3 in default non-btree mode, aren't.
> I doubt UFS is any better).
I hear that often but don't give it much credence. We use ext3 in a
Murder with 70K accounts and two backen
On Wed, 25 Feb 2004, Etienne Goyer wrote:
> On Wed, Feb 25, 2004 at 01:11:30AM -0300, Henrique de Moraes Holschuh wrote:
> > 3. A proper filesystem (ext2 and ext3 in default non-btree mode, aren't.
> > I doubt UFS is any better).
>
> I hear that often but don't give it much credence. We us
OTECTED]
Henrique de Moraes Holschuh <[EMAIL PROTECTED]>
Sent by: [EMAIL PROTECTED]
25-02-2004 00:11
To
Michael J Barber <[EMAIL PROTECTED]>
cc
[EMAIL PROTECTED]
Subject
Re: Performance question...
On Tue, 24 Feb 2004, Michael J Barber wrote:
> Cyrus after he thought it was too
sizes too big. I switched to DIGEST-MD5 and this problem stopped. Any
info you can give me would be great.
- Original Message -
From: "Etienne Goyer" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Wednesday, February 25, 2004 11:03 AM
Subject: Re: Performance
On Wed, Feb 25, 2004 at 02:59:01PM -0400, Cesar Lagarrigue wrote:
> > 1. noatime
> how configure this ?
In /etc/fstab, specify the "noatime" option for the partition on which
your mailspool reside.
> > 2. no "sync mounts" or "sync attributes" anywhere. Use a proper fs
> >instead.
> i use
^^
^^ how configure this ? is in the imapd.conf ?
^^
^^ p.d: :) thanks for any helps.
^^
^^ Regards,
^^
^^ Is you need send any conf please tell me.
^^
^^
^^ Cesar Lagarrigue V.
^^ Project Engineer
^^ [EMAIL PROTECTED]
^^
^^
^^
^^
^^ Henrique de Moraes Holschuh <[EMAIL PROTECTED]>
On Wed, 25 Feb 2004, Etienne Goyer wrote:
> > > 4. A properly configured Berkeley DB environment (although for some
> > > reason, Cyrus seems to actually survive well without any config,
> > > maybe the CMU crew override the defaults with something sane in
> > > the bdb backend code
On Wed, Feb 25, 2004 at 04:22:48PM -0300, Henrique de Moraes Holschuh wrote:
> On Wed, 25 Feb 2004, Etienne Goyer wrote:
> > > > 4. A properly configured Berkeley DB environment (although for some
> > > > reason, Cyrus seems to actually survive well without any config,
> > > > maybe the
Etienne Goyer wrote:
First, you need DB4.2 if you have anything SMP or multithreaded :)
Ok, this is getting interesting. Are there known with db3 on SMP
system? We use the stock db3 rpm from RedHat 7.3 (3.3.11) on an SMP
machine, and we seem to have database corruption problem on
mailboxes.db.
On Wed, 25 Feb 2004, Henrique de Moraes Holschuh wrote:
> First, you need DB4.2 if you have anything SMP or multithreaded :)
Ummm, we've successfully used Berkeley DB on SMP systems since DB3... Is
there a reference to what problems this causes anywhere?
-Rob
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
On Wed, 25 Feb 2004, Etienne Goyer wrote:
> On Wed, Feb 25, 2004 at 04:22:48PM -0300, Henrique de Moraes Holschuh wrote:
> > On Wed, 25 Feb 2004, Etienne Goyer wrote:
> > > > > 4. A properly configured Berkeley DB environment (although for some
> > > > > reason, Cyrus seems to actually survi
On Wed, 25 Feb 2004, Rob Siemborski wrote:
> On Wed, 25 Feb 2004, Henrique de Moraes Holschuh wrote:
> > First, you need DB4.2 if you have anything SMP or multithreaded :)
>
> Ummm, we've successfully used Berkeley DB on SMP systems since DB3... Is
> there a reference to what problems this causes
On Wed, 25 Feb 2004, Henrique de Moraes Holschuh wrote:
> db 3.2 is safe. I don't know about 3.3. 4.0 and 4.1 are a NO GO for SMP.
Is this all Linux SMP installations, or just those using NPTL?
--
Simon Brady mailto:[EMAIL PROTECTED]
ITS Technical Services
University
-- Henrique de Moraes Holschuh <[EMAIL PROTECTED]> is rumored to have mumbled
on Mittwoch, 25. Februar 2004 17:39 Uhr -0300 regarding Re: Performance
question...:
On Wed, 25 Feb 2004, Rob Siemborski wrote:
On Wed, 25 Feb 2004, Henrique de Moraes Holschuh wrote:
> First, you need DB4
[EMAIL PROTECTED] root]# rpm -qi db4
Name: db4 Relocations: (not relocateable)
Version : 4.1.25Vendor: whiteboxlinux.org
Release : 8 Build Date: Fri 12 Dec 2003
02:54:32 AM CLST
Install Date: Wed 11 Fe
PROTECTED]
Henrique de Moraes Holschuh <[EMAIL PROTECTED]>
Sent by: [EMAIL PROTECTED]
25-02-2004 15:22
To
Etienne Goyer <[EMAIL PROTECTED]>
cc
[EMAIL PROTECTED]
Subject
Re: Performance question...
On Wed, 25 Feb 2004, Etienne Goyer wrote:
> > > 4. A properly con
Sebastian Hagedorn wrote:
-- Henrique de Moraes Holschuh <[EMAIL PROTECTED]> is rumored to have
mumbled on Mittwoch, 25. Februar 2004 17:39 Uhr -0300 regarding Re:
Performance question...:
On Wed, 25 Feb 2004, Rob Siemborski wrote:
On Wed, 25 Feb 2004, Henrique de Moraes Holschuh
Greetings,
Ok, time for another question... We are running a mail system (ISP) with just a bit
over 4000 mailboxes at present... Our exact configuration is as follows: RedHat
Linux 7.0 + patches, postfix snapshot 20010204 (MTA), and cyrus-imapd 2.0.11 system
is a dual PIII 800 with 512mb r
f hurt in a week or two :-) Let me know how it
goes, I'm anxious to find out.
Cheers
Chris
- Original Message -
From: "Andy Hubbell, Jr." <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Friday, February 09, 2001 11:33 PM
Subject: Performance Question cyrus-i
> On Sat, 10 Feb 2001 12:57:35 -0500,
> Christopher Audley <[EMAIL PROTECTED]> (ca) writes:
ca> The LMTP update for postfix is found on the download page at www.postfix.org
ca> under "Unofficial patches". I don't know how well it works. Alternately,
Oh my, this is way out of date. Sim
world of hurt in a week or two :-) Let me know how it
>goes, I'm anxious to find out.
>
>Cheers
>Chris
>
>- Original Message -
>From: "Andy Hubbell, Jr." <[EMAIL PROTECTED]>
>To: <[EMAIL PROTECTED]>
>Sent: Friday, February 09, 2001 11
On Sat, Feb 10, 2001 at 12:57:35PM -0500, Christopher Audley wrote:
> The LMTP update for postfix is found on the download page at www.postfix.org
> under "Unofficial patches". I don't know how well it works.
We are using snapshot 2000-12-17 which has LMTP support and had no problems
so far.
G
> The LMTP update for postfix is found on the download page at
> www.postfix.org under "Unofficial patches". I don't know how well it
> works. Alternately, if you want to go with production versions of software,
> sendmail has native support for LMTP in the recent releases. You could move
> the
On 10-Feb-01 at 14:07, Andy Hubbell, Jr. ([EMAIL PROTECTED]) wrote:
> Christopher,
>
> Sorry to inform you, but since I am using a very recent snapshot of postfix
> it does support lmtp "out of the box" as it were. And as I stated in my
> email, I am using lmtp directly via the socket interface.
> On Sat, 10 Feb 2001 15:54:40 -0500,
> Andy Hubbell, <[EMAIL PROTECTED]> (ah) writes:
ah> Christopher,
ah> Sorry to inform you, but since I am using a very recent snapshot
ah> of postfix it does support lmtp "out of the box" as it were.
ah> And as I stated in my email, I am using lmtp di
Amos,
Aha! The vital piece that I was missing! ;) I had set local_transport
and mailbox_transport and even fallback_transport in my attempts to get
this straight... So I'm assuming for my local domains I need to just put
the entry in a transport map for my local domains and it will use the
Pat,
I'm not seeing any evidence that postfix is batching up the local
deliveries through a single lmtp instance... I've made some changed
recommended by others using postfix, but still don't see evidence of this
happening... Not sure if its my postfix config or my cyrus config that's
causi
On 11-Feb-01 at 11:39, Andy Hubbell, Jr. ([EMAIL PROTECTED]) wrote:
> I'm not seeing any evidence that postfix is batching up the local
> deliveries through a single lmtp instance... I've made some changed
> recommended by others using postfix, but still don't see evidence of this
> happening.
> On Sun, 11 Feb 2001 12:44:32 -0500,
> Andy Hubbell, <[EMAIL PROTECTED]> (ah) writes:
ah> Should the master process exec a new lmtpd for each delivery? This
ah> would seem somewhat wasteful, especially on a busy system with lots
ah> of disk IO... Nothing else is talking lmtp to this se
ions are not too likely to come in near future
--
Bruno
- Original Message -
From: Amos Gouaux <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Sunday, February 11, 2001 9:15 AM
Subject: Re: Performance Question cyrus-imapd 2.0.11
> >>>>> On Sat, 10 F
Another thing I forgot to mention - you might try linking lmtpd
static to reduce per-instance startup costs. (This may not be
an issue if setting prefork non-zero causes the lmtp daemon to
fork for each instance instead of the master forking and then
execing lmtpd.)
-Pat
On 11-Feb-01 at 15:23, Bruno Riedl ([EMAIL PROTECTED]) wrote:
> You will run into new problems using a transport map for local lmtp
> delivery. From my understanding, Postfix only evaluates aliases when
> doing local delivery. A transport map is no such thing, so You won't
> have Your aliases ex
> On Sun, 11 Feb 2001 22:45:46 +0100,
> Bruno Riedl <[EMAIL PROTECTED]> (br) writes:
br> From my understanding, Postfix only evaluates aliases when doing local
br> delivery.
br> A transport map is no such thing, so You won't have Your aliases expanded.
All you have to do is use a virtual
> On Sun, 11 Feb 2001 16:58:04 -0800 (PST),
> patl <[EMAIL PROTECTED]> (p) writes:
p> If this is true, you might want to look into replacing postfix with
p> exim. Exim is very easy to configure to handle this sort of thing.
For us it hasn't really been a problem. We seem to be somewha
42 matches
Mail list logo
