March 9, 2006 at 2:25 pm
Hi All,
I am new to this forum, and I am hoping that you good folks here can provide some insight about this suggested server configuration for High Availability.
Essentially, I have an Intranet application for a mid-size organization (about 200 employees). Based on the nature of the application, even if everyone bangs our server at the same time, I don't imagine that our DB load can't be handled by a 2-way, dual-core system. However, what I am worried about is hardware failure. The application (or, more precisely, its users) cannot tolerate more than a handful of minutes of downtime or data loss. Based on that, here is my thought for our hardware config:
Primary DB Server - 2xdual core pentiums or opterons, with oodles of Ram and disk space, SQL Enterprise 2K (we're not migrating to 05 yet).
Secondary DB Server- 1x dual core pentium or opteron, with less ram and disk space than the primary, SQL Standard 2K.
Either or both of these might be connected to a SAN unit
Essentially, I want to log ship from the primary to the secondary. This will help me in two ways:
1) When the primary is active, the secondary will simply get regular updates with log shipments. In addition, the secondary can also be used for processing offline reports so that they don't eat cpu cycles from the primary.
2) When the primary goes down, I can point my applications with minimal changes to the secondary and promote it in a 'Warm' backup mode. While I know that this will have diminished capacity, I also know that it is a temporary fix.
I am curious if people have any comments or suggestions on how to make this better or why it is no good.
Thanks,
--Yonah
I am Doing it with .Net, are you?
March 13, 2006 at 8:00 am
This was removed by the editor as SPAM
March 13, 2006 at 10:57 pm
March 14, 2006 at 7:40 am
Trigger,
The idea here is this, AFAIK, the investment I would need to make in hardware and software to setup a clustered environment would be significantly more expensive than losing, say, 15 minutes of data and 30 minutes of uptime.
I am Doing it with .Net, are you?
March 20, 2006 at 2:24 pm
If your users cannot tolerate more than a handful of minutes. Lomg ship is not a solution. I would consider the followings:
1. SQL Cluster locally (with disk mirroing etc) for High Availability.
2. Transactional Replication to a remote site for DR.
Good luck!!
Viewing 5 posts - 1 through 4 (of 4 total)
You must be logged in to reply to this topic. Login to reply