Merge Replication between sites - Keeping hot backu servers ready

Dear all
I have set a merge replication between servers say SERVER A (Publisher) and
SERVER B (Subscriber to A)
The agent runs every 2 minutes.
I have set all identity columns with increment value by 2. SERVER A has all
odd identity values and SERVER B has even.
I have 2 more servers (STAND BY) say SERVER C and SERVER D.
SERVER C is located near to SERVER A (Same office - Office A) and SERVER D
is near SERVER B (Same Office - Office B).
I want to keep SERVER C as backup replica (in case SERVER A goes down users
will connect to it and there will be minimum downtime for the users in
Office A and when SERVER A comes up it will synch with SERVER C and in turn
it will synch with SERVER B) of SERVER A (which is publisher to SERVER B) so
this can be publisher to SERVER C.
How do I configure SERVER D which will be replica (in case SERVER B goes
down users will connect to it and there will be minimum downtime for the
users in Office B and when SERVER B comes up it will synch with SERVER D and
in turn it will synch with SERVER A) of SERVER B.
Will it be subscriber or publisher to SERVER B (Which is subscriber to
SERVER A).
1- Do I need to configure SERVER B as Publisher and Subscriber?
2- Do I need to configure SERVER D as Publisher and SERVER B as subscriber.
In this case will it replicate between SERVER A and SERVER B. Do I need to
configure SERVER A's subscribers can synch with multiple PUBLISHER?
Your inputs are highly appreciated.
Hope I conveyed my need clearly.
Regards
Javed Iqbal
Javed,
you could set ServerB up as a republisher - subscriber to A and publisher to
D. However I think an easier solution would be to have B, C and D as
subscribers to A if your network speeds can cope with it. In this case, if B
goes down, users use D which is regularly synchronized with A. When it comes
back on line, A and B are synchronized.
Cheers,
Paul Ibison SQL Server MVP, www.replicationanswers.com
(recommended sql server 2000 replication book:
http://www.nwsu.com/0974973602p.html)
|||Thanks Paul,
Initially I thought about same keeping A as publisher and rest all
subscribers. But the connectivity between SERVER A and SERVER B (2 MBPS) may
not be enough. There will be lot of other network traffic between office and
can hamper the replication process.
Thank you once again for your prompt help.
Regards
Javed Iqbal
"Paul Ibison" <Paul.Ibison@.Pygmalion.Com> wrote in message
news:O1$4JwMyFHA.2848@.TK2MSFTNGP15.phx.gbl...
> Javed,
> you could set ServerB up as a republisher - subscriber to A and publisher
to
> D. However I think an easier solution would be to have B, C and D as
> subscribers to A if your network speeds can cope with it. In this case, if
B
> goes down, users use D which is regularly synchronized with A. When it
comes
> back on line, A and B are synchronized.
> Cheers,
> Paul Ibison SQL Server MVP, www.replicationanswers.com
> (recommended sql server 2000 replication book:
> http://www.nwsu.com/0974973602p.html)
>
|||Using merge replication for disaster recovery is a bad choice for the
following reasons
1) you add a rowguid column to each table you are replicating
2) increased latency of transactions occurring on the publisher/subscriber
due to the triggers
3) long sync times
4) no automatic recovery
5) somewhat limited abilities to modify schema
6) possibility of data loss
For automatic failover you need to use clustering. Otherwise a better
solution is probably log shipping.
Hilary Cotter
Looking for a SQL Server replication book?
http://www.nwsu.com/0974973602.html
Looking for a FAQ on Indexing Services/SQL FTS
http://www.indexserverfaq.com
"Javed" <javediqbal98@.hotmail.com> wrote in message
news:uRHPgnMyFHA.788@.tk2msftngp13.phx.gbl...
> Dear all
> I have set a merge replication between servers say SERVER A (Publisher)
and
> SERVER B (Subscriber to A)
> The agent runs every 2 minutes.
> I have set all identity columns with increment value by 2. SERVER A has
all
> odd identity values and SERVER B has even.
> I have 2 more servers (STAND BY) say SERVER C and SERVER D.
> SERVER C is located near to SERVER A (Same office - Office A) and SERVER D
> is near SERVER B (Same Office - Office B).
> I want to keep SERVER C as backup replica (in case SERVER A goes down
users
> will connect to it and there will be minimum downtime for the users in
> Office A and when SERVER A comes up it will synch with SERVER C and in
turn
> it will synch with SERVER B) of SERVER A (which is publisher to SERVER B)
so
> this can be publisher to SERVER C.
> How do I configure SERVER D which will be replica (in case SERVER B goes
> down users will connect to it and there will be minimum downtime for the
> users in Office B and when SERVER B comes up it will synch with SERVER D
and
> in turn it will synch with SERVER A) of SERVER B.
> Will it be subscriber or publisher to SERVER B (Which is subscriber to
> SERVER A).
> 1- Do I need to configure SERVER B as Publisher and Subscriber?
> 2- Do I need to configure SERVER D as Publisher and SERVER B as
subscriber.
> In this case will it replicate between SERVER A and SERVER B. Do I need to
> configure SERVER A's subscribers can synch with multiple PUBLISHER?
> Your inputs are highly appreciated.
> Hope I conveyed my need clearly.
> Regards
> Javed Iqbal
>