My question is this: say, for whatever
[quoted text, click to view] >reason, one of the sites misses out on a transaction
update. will it
>go out of sync? If so, how does replication get back up
to date? Does
>it automatically re-initialize itself, and then get
picked up by the
>snapshot agent? Or will I need to panic and do it myself?
>
From my limited experience, replication is able to catch
it self back up. Say during the transfer of lots of data
from the Pub to the Sub the line goes down. Replication is
smart enough to rollback what data had already made it to
the Sub. Then when the line is back up, it resends it. I
would suggest lots of testing for scenarios like this to
see for yourself what would happen and what you would need
to do in time of disaster.
[quoted text, click to view] >-----Original Message-----
>Hi
>
>We have a couple of sites across a WAN that need access
to live data.
>There are a few access databases set up to link to this
data at source
>- but the problem with access is that it pulls all the
data across the
>line to process it locally. This is very slow for some of
our further
>away sites (probably not a very quick WAN either).
>
>After some consideration, I have set up transactional
replication to
>the sites in question, and re-linked their copies of the
access
>databases to point to the local data.
>
>It all seems to work ok. My question is this: say, for
whatever
>reason, one of the sites misses out on a transaction
update. will it
>go out of sync? If so, how does replication get back up
to date? Does
>it automatically re-initialize itself, and then get
picked up by the
>snapshot agent? Or will I need to panic and do it myself?
>
>Thanks,
>Mark.
>.