Gail,

Cheers, thats the explanation i needed.

I imagine using bulk logged on the subscription replica is a replication throughput issue,

all the insert update action from the replication stored procs will be getting logged in the simple...

Perry,

Your article suggested you could make the single Listener HA, Could you explain how this is achieved please ?

my target table is 44Million rows and has an int PK but is clustered on the Site and the PK with the site First. I typically merge 107k...

Incidentally there is also a trace flag you can use to stop the exclusive table lock escalation kicking in at 5000 Rows (try Books on line for it)

another option for...

not tried your method but as cdc is pretty new to sql id prob stick with an older method:

triggers on change tracked tables that write change data to seperate local...

I use the Following, in our scenario we are merging a site into a table that contains all sites i find that the CTE restricts the query plan on the...

I dont think the trigger solution is what was asked for. ie (monitoring structual changes)

It seemed to be more about the realease process for making structural changes to the DB...

You said this was a histric problem. have you looked at and retained the default trace. it may tell you what was going on at the time of the...

Hi,

what sticks out in this scenario to me is the large number of databases on a single instance! SQL will be constantly re-allocating memory to the active DB at that...

yep you just restore the latest diff backup

remember all your difs work from the baseline of the Last full backup Only and dont contain the difs from the last diff. so they get progressivly Larger during...

no problem, youve illustrated a fair point. Id like to see the link that adds some context to your example

good news,

the unistall of sp1 seemed to work and 3 hours later my sql services are now starting up correctly again.

unfortnualty this confirms that it is definatly the Vista SP1...

clever but on large tables probably inparactical due to lock escalation and the time taken to do disk based delete and reindex operations.

alternatives could be a Unique indexed view ...