Distribution server restore on standby server

Hi,
we have developed an application using transactional replication, with
several publication servers (push) and a distribution server (which is
the only subscriber too). It seems to work fine.
Now we would "protect" the distribution/subscriber server by using a
standby server (note: only the distribution/subscriver server must be
protected, not publication servers) which should replace the working
server in the case it crashes.
Could someone suggest us the best strategy to do this? Thanks in
advance...
Marco
You can manually set this up. Have a look at Strategies for Backing Up and
Restoring Transactional Replication in BOL.
The problem is that this adds to the latency. Transactions remain in your
tlog until the log is dumped. Then they are read from the tlog and written
to the distirbution database.
"Marco69" <marcosindona@.virgilio.it> wrote in message
news:ce7beb14.0403250541.5f374a8f@.posting.google.c om...
> Hi,
> we have developed an application using transactional replication, with
> several publication servers (push) and a distribution server (which is
> the only subscriber too). It seems to work fine.
> Now we would "protect" the distribution/subscriber server by using a
> standby server (note: only the distribution/subscriver server must be
> protected, not publication servers) which should replace the working
> server in the case it crashes.
> Could someone suggest us the best strategy to do this? Thanks in
> advance...
> Marco