[NTISP Digest]

ntisp-digest-request@iea-software.com
Sun, 12 Jul 1998 00:00:26 -0700

Message 1: Re: Redundancy & Fault Tolerance
from "Mike Stankavich" <mike@ethergate.com>

..------ ------ ------ ------ ------ ------ ------ ------ ------ ------.
| Message 1 |
'------ ------ ------ ------ ------ ------ ------ ------ ------ ------'
Subject: Re: Redundancy & Fault Tolerance
From: "Mike Stankavich" <mike@ethergate.com>
Date: Sat, 11 Jul 1998 10:57:10 -0700

Victor,

I tried Octopus by Qualix Technologies (http://www.octopustech.com) but I
wasn't impressed. It stopped replicating a lot and would then fill up the
hard disk with the replication log. Also you can fail over a maximum of 25
IP addresses. Doesn't do much good when you are running 60 virtual domains
that each have an IP address. Vinca has a solution that looks pretty good
(http://www.vinca.com), but I haven't tried it. I'd love to hear from
anybody that has. Then there's Microsoft Cluster Server which looks harder
to implement & a fair bit more expensive. I really haven't looked into MSCS
(wolfpack) much at all. I'll be interested to see how it works out for you.

Mike Stankavich, CTO

NorFab/ISP, Inc.
921 SW Washington St
Portland OR 97205

mike@ethergate.com

(503) 972-2822 direct
(503) 972-2810 office

-----Original Message-----
From: Victor A. <victor@webjogger.net>
To: ntisp@iea-software.com <ntisp@iea-software.com>
Date: Friday, July 10, 1998 4:29 PM
Subject: Redundancy & Fault Tolerance

I need the expert advice of those of you that have implemented some sort of
Redundancy and fault tolerance in your location. Right now I only have a
single Web Server running IIS 4.0, SQL 6.5, CF 3.1 and a couple of other
applications that run on top of the Web Server. I do back everything every
night but that's not what I'm interested in at this point. I'm adding a new
server next week and I want to implement redundancy in two ways (please
correct me if I'm dreaming here):

1) I want the load be distributed in both servers and

2) If one of the servers fails I want the other one to immediately take
over so that not a single user notice our Web Services are down.

I know this can be done, the question is: What is the best way to
accomplish it?

If anyone out there has done something similar and is willing to share the
"know how" I will greatly appreciate it.

Best to all,
-Victor