April 25, 2006 at 10:26 pm
Comments posted to this topic are about the content posted at http://www.sqlservercentral.com/columnists/mcook/achievingserverredundancyatremoteoffices.asp
May 25, 2006 at 6:41 am
I'm not sure I follow these IP addresses:
Production Server: Name : PROD1 IP : XXX.XX.XX.223
Secondary Server: Name : LOGSHIP1 IP : XXX.XX.XX.224
We create an alias like this:
Alias : Name : PROD1V IP : XXX.XX.XX.223
shouldn't there be 3 total?
May 25, 2006 at 6:51 am
It depends. sounds like a microsoft answer doesn't it. But seriously, if you include the unused IP address the failed production server gets changed to then yes - 3 IP addresses will be needed. If you're only talking about normal production operation, then you only need two - one for the alias and the production server and another for the log shipping server.
May 25, 2006 at 7:12 am
Great article. - Tim Chapman
May 25, 2006 at 7:59 am
Good article.
By the way, it sounds like you've got a couple of unused licenses.
You do not need to buy a sql server or windows server license for the passive server - either in log shipping or clustering. Unless of course you're using the SQL Server for other, non-passive, purposes.
Taken from Microsoft's licensing white paper:
"When doing failover support, a server is designated as the passive server. The purpose of the passive server is to absorb the data and information held in another server that fails. A passive server does not need a license, provided that the number of processors in the passive server is equal or less than those of the active server. The passive server can take the duties of the active server for 30 days. Afterward, it must be licensed accordingly."
May 25, 2006 at 8:02 am
thanks for the information. I knew only one license was needed when using Microsoft Log Shipping included in Enterprise edition, but thought that anything outside that required additional licenses. Again, thanks...
May 25, 2006 at 8:36 am
You're probably better off maintaining the second license if you go ahead with your strategy of having the secondary work for multiple primaries. It might be in use frequently, albeit for short periods each time.
But... if you have multiple primaries for the secondary, I don't think your switch of the secondary's IP address and alias is going to work as well. What if two primaries fail at once? It's unlikely, I'm sure, and this article is a good treatment of a reasonable, low-cost and responsible method. Thanks!
May 25, 2006 at 8:41 am
How are you handling network outages?
K. Brian Kelley
@kbriankelley
May 25, 2006 at 8:45 am
thanks for the comments. you're correct, the setup can accommodate only a single failure in the multiple primary-single secondary situation, but our assessment was that you can go pretty far down the proverbial rabbit hole in trying to achieve redundancy, and that this was as you said a reasonable, low-cost solution....
May 25, 2006 at 9:03 am
the two servers - primary and secondary are located next to each other on the local lan. In the event that there is a problem with the primary and we need to failover to the secondary ... and the wan link is down, there are procedures in place that a designated person at the facility can be guided through by a DBA....or did i miss your point entirely...
May 25, 2006 at 9:37 am
and the wan link is down, there are procedures in place that a designated person at the facility can be guided through by a DBA
Nope, you got it. That's what I was looking for. I didn't see that in your article and this is an area some organizations miss.
K. Brian Kelley
@kbriankelley
May 25, 2006 at 9:43 am
thanks for the comments so far, all are appreciated and noted...
May 25, 2006 at 11:04 am
A question please.
When failure occurs in the primary server, how can you get data from the last synchronization to the time before the failure?
May 25, 2006 at 12:16 pm
depending on the availability of the server, you may be able to ship the last log, etc. if the server is completely unavailable, there's not much you can do except accept that some data will be lost. what we tried to do was to gauge how much time we could afford to lose in the event of a system failure. obviously you can configure longer or shorter time intervals, but the less time you can afford to lose, the more you need to look at something like clustering or perhaps database mirroring in SQL2K5 that provides more real-time synchronization.
May 30, 2006 at 6:58 am
Did you consider replication as an option?
Viewing 15 posts - 1 through 15 (of 19 total)
You must be logged in to reply to this topic. Login to reply