General Replication question

Question

General Replication question

Animal Magic

SSChampion

Points: 13964
More actions
September 3, 2007 at 3:22 am

#98491

If you are using transactional replication from server A to server B and server A goes down, you failover to server B to continue uptime. What would you generally do to get server A back up to speed? would you replicate back? or would you wait until there is server downtime and do a backup and restore? Your views/experiences would be appreciated
Thanks
John

Viewing 10 posts - 1 through 9 (of 9 total)

You must be logged in to reply to this topic. Login to reply

Paul Mu SSCertifiable Points: 5547 More actions · Answer 1

For high availability, you would use SQL Server Clustering. For disaster recovery go for things like log shipping and database mirroring in SS 2005.

Replication is often used for scalibility reasons as in reporting functions.

HTH

Paul

Erik Bitemo SSCrazy Points: 2062 More actions · Answer 2

Agree with Paul. You can use also log shipping for business continuity, with which you can quite easily bring your lost database up-to-date. Read this series: http://www.microsoft.com/technet/prodtechnol/sql/2000/deploy/sqlhalp.mspx

-- Erik http://blog.rollback.hu

Graeme100 SSCrazy Eights Points: 8054 More actions · Answer 3

Well it does, of course, depend on the reason for the replication in the first place. If this is for the purpose of DR then there are better ways of doing this as has been mentioned. If it is for the prupose of transferring data to a reporting server and this server is more or less identical to the failed server and there are no other processes dependant on server A...then server B could become your new server A and you can build a new server B....if that makes sense.

There many ifs and buts here so the answer can only be a general one.

Apologies for that

Regards

Graeme

Animal Magic SSChampion Points: 13964 More actions · Answer 4

Hi Graeme,

Thanks for the response. The reason for the "serverB" is reporting and like you say if serverA does encounter problems bringing serverB online isnt an issue. My question was (apart from badly worded) what methods would you use to get the new server (serverA rebuild) back up to speed?

noeld SSC Guru Points: 96590 More actions · Answer 5

You simply set replication back. Now, there are *many* tricks in which your environment could handle it and those are very dependent on your availability.

The plan on some places is to simply backup, restore and set replication back without initial snapshot. On some others is just to create the snapshot and apply it. On some other more special cases you add table articles little by little to the publication and run periodic snapshots until fully replicated.

Again there are many twists and you have to use the ones that serve you best.

* Noel

Animal Magic SSChampion Points: 13964 More actions · Answer 6

Thanks Noel, i guess i will just use a bit of trial and error (although hopefully the main server wont go down too much!!). I'm thinking as i have plenty of downtime at night and a relatively small database a backup and restore would probably do the trick.

Cheers

Erik Bitemo SSCrazy Points: 2062 More actions · Answer 7

John,

Aren't you going to upgrade to SQL 2005? There you can use peer-to-peer (bidirectional multimaster) replication which can give you the required result.

-- Erik http://blog.rollback.hu

Animal Magic SSChampion Points: 13964 More actions · Answer 8

Yes we are planning on moving to 2005 before xmas, although i have advised the company (i only started here last week) that we should wait for 2k8 in Feb. the peer to peer replication sounds like something looking into tho when we do move! Cheers

JeremyE SSCoach Points: 16201 More actions · Answer 9

You may be disappointed if you believe SQL 2008 will be available in Feb. 2008.

Microsoft says the product will be "Released" in Feb. 2008 but "Available" and "Released" are 2 different things to Microsoft.