[Linux-HA] High availability, fast, fileservers
david.lang at digitalinsight.com
Sat Mar 19 23:35:18 MST 2005
you may want to take a look at the cyrus mail server software, they have a
cluster layer (called Murder) that will effectivly spread things across a
large number of servers. no need for NFS (and it's inherent locking
On Sat, 12 Mar 2005, John R.Shearer wrote:
> Date: Sat, 12 Mar 2005 08:40:46 -0800
> From: John R.Shearer <john at puremail.com>
> Reply-To: General Linux-HA mailing list <linux-ha at lists.linux-ha.org>
> To: Jeff Tucker <linux-ha at lists.linux-ha.org>
> Subject: [Linux-HA] High availability, fast, fileservers
> I was reading a thread you started almost exactly a year ago on the Linux
> High Availability mailing list
> You were running a large Email service environment and wanted scalability and
> high availability. The issues discussed in the thread are exactly the issues
> I am grappling with. I wanted to know how your system turned out.
> I am developing the specifications for a 100,000 mailbox Email cluster where
> each user receives 1 GB of space (100 TB total capacity). It was my
> intention to use 14 NFS servers, each with a belly full (24) of 400 GB SATA
> disks divided among 2 SATA controllers and RAID 5'd, and use DRBD to mirror
> them to another 14 NFS servers. Each of my 6 or so mail and web servers
> would then NFS mount all 28 NFS shares from the 14 active NFS servers.
> What did you end up using for your environment? What is your current
> capacity? Also, for my estimates, what are the averages I can expect for A)
> number of messages received per mailbox per day, and B) the average size of
> an Email message.
> Thanks for your time. - John R. S.
> Linux-HA mailing list
> Linux-HA at lists.linux-ha.org
There are two ways of constructing a software design. One way is to make it so simple that there are obviously no deficiencies. And the other way is to make it so complicated that there are no obvious deficiencies.
-- C.A.R. Hoare
More information about the Linux-HA