Please start any new threads on our new
site at https://forums.sqlteam.com. We've got lots of great SQL Server
experts to answer whatever question you can come up with.
Author |
Topic |
beckkp
Starting Member
3 Posts |
Posted - 2004-08-18 : 22:37:57
|
Hi,Im currently implementing an application that may need 2 SQL2000 running at the same time...In case that one is down, another one will be up immediately to replace it...Im still finding the solution for this...have considered the Replication but I think it wouldn't automatic switch the server, which is down, to another...Anyone can give me suggestions or may be solutions...thanks !!!==========||beckkp||========== |
|
derrickleggett
Pointy Haired Yak DBA
4184 Posts |
Posted - 2004-08-19 : 08:44:42
|
You have a few options here:1. Active/passive cluster--This provides you with server redundancy, not "drive" redundancy. If one server fails though the other picks up.--The applications need to be cluster aware.--Cheapest to set up. You don't have to pay for SQL licensing on passive server unless you are doing something else on it.2. Log Shipping.--You have to pay for SQL Server licensing on both servers.--It's not "immediate" failover. You would have a few minutes of downtime in a dr situation.3. Replication.--You have to pay for SQL Server licensing on both servers.--It's not "immediate" failover. You would have a few minutes of downtime in a dr situation.--Generally faster failover than log shipping.--Harder to maintain.4. Active/Active Cluster with replication.--Expensive and more maintenance than any of the above.--Provides server reduncdancy mixed with drive redundancy.--Very complex.5. Cold Site Redundancy.--Use log shipping to ship to remote site.--Slower recovery than above.--Covers facility loss.6. Hot Site Redundancy.--Most expensive solution.--Many times requires rewriting application or hosting different apps in different location.--Involves expensive hardware and software solutions from providers that specialize in this.--Need enough bandwidth to pull it off.--Most redundant of all solutions if implemented correctly.MeanOldDBAderrickleggett@hotmail.comWhen life gives you a lemon, fire the DBA. |
|
|
Kristen
Test
22859 Posts |
Posted - 2004-08-19 : 12:59:16
|
V. useful - chuck it in your blog??Kristen |
|
|
beckkp
Starting Member
3 Posts |
Posted - 2004-08-27 : 05:35:46
|
thanks..derrickleggett...Btw, I have tried the solution wif MSCS on win2003 Enterprise..since I have only 2 nodes, I use Majority Node Set but the problem is that SQL2000 SP1 (will be upgrade to SP3 after installed) need a shared disk drive between 2 node to set up the virtual server.....I have been struggling one whole day.....Im wondering if we can use any other ways rather than the SCSI and FibreChannel for shared array disks...some kind of network sharing drive...!!??Thanks much!!!!==========||beckkp||========== |
|
|
MuadDBA
628 Posts |
Posted - 2004-08-27 : 10:29:41
|
In this case, there is no substitue for an experienced DBA with expertise in DR configuration.Tell your manager you need to hire a consultant for this....it's the best favor you could do him, and yourself. |
|
|
|
|
|
|
|