D
dbeavon
Guest
>> you don't have to replicate *everything*. Hi Tom. Isn't it the same with DataServer? You don't have to put everything in DataServer for SQL either. The only real concern I can think of is to ensure that *all* of the tables that participate in a given transaction are on one side or the other. (ie. your update operations should hit the transaction log of one or the other, but not span across both). This isn't an unreasonable restriction. Most of our "master table" reads are NOLOCK anyway. (For those tables it really doesn't even matter where the data lives, given that the records are not honoring or participating in the ACID requirements of our transactions). For the tables we *do* choose to keep in SQL, it seems that both technologies require that we maintain the SQL schema *and* the "schema holder", right? And for PRO2SQL there is an additional schema concern that involves keeping the Progress schema synchronized with SQL Server (.. but DataServer doesn't care about this since it doesn't have that additional schema). Where PRO2SQL is concerned, are any of these schema-related concerns automated away - to the point where OE developers would have *less* effort than they would with DataServer? It seems to me that the presence of an additional database would always create *more* development work, not less.
Continue reading...
Continue reading...