Subject | Re: [firebird-support] Firebird & Compatibility idea |
---|---|
Author | Helen Borrie |
Post date | 2006-10-27T08:53:51Z |
At 05:43 PM 27/10/2006, you wrote:
accessing both.
writes its changes to a replication table. The replication process
on the main server periodically connects to both the source and the
target database, reads its replication table and passes requests to
the target database. In most replication systems, those requests are
passed as parameters to stored procedures in the target database,
which has the replication rules installed.
the clients or in some intermediate layer. In a distributed
environment, there would be network-wide server management software
taking care of such things.
means. You would also need to have a replication server on the
standby server, to ensure that changes in the second database are
replicated back to the main database when it comes back on-line; and
something programmed in your recovery procedures to ensure that the
two databases are synchronised before normal operations resume.
and forms. There are plenty of Firebird servers out there that are
participants in "clustered" environments.
firebird-enterprise list and ask there about environments that people
are running Firebird in....
./heLen
>Dear Sir,Not the *same* database, but two databases, with one or both servers
>
>We want to use a main server and a secondary server, both would need to
>access to the same database.
accessing both.
>The idea is : the main server running and processing the request to theFor this model, your main server runs and processes the requests and
>database while the second server is replicating them.
writes its changes to a replication table. The replication process
on the main server periodically connects to both the source and the
target database, reads its replication table and passes requests to
the target database. In most replication systems, those requests are
passed as parameters to stored procedures in the target database,
which has the replication rules installed.
>If for a reason or an other, the main server is out of service,"Automatically" will be something you program or script externally on
>burning or anything else, the second server should take all the works,
>and this automatically.
the clients or in some intermediate layer. In a distributed
environment, there would be network-wide server management software
taking care of such things.
>If main server is back, he takes back the works. (also automatically)Again, this type of automation would be achieved by external
means. You would also need to have a replication server on the
standby server, to ensure that changes in the second database are
replicated back to the main database when it comes back on-line; and
something programmed in your recovery procedures to ensure that the
two databases are synchronised before normal operations resume.
>We looked at Cluster server but it seems Firebird doesn t works withBeware of banal buzzwords!! Clustering solutions come in many shapes
>Cluster.
and forms. There are plenty of Firebird servers out there that are
participants in "clustered" environments.
>Could you confirm me this and at least, give us an other solution toOh boy, that's a loaded question! You might like to subscribe to the
>be able to give to our customer a unbreakable service.
firebird-enterprise list and ask there about environments that people
are running Firebird in....
./heLen