Hello all,
Never posted here before, so I apologize if I make some errors in posting.
We use Replication, but it is my belief we do not use it smartly. The person that set it up did an excellent job of doing so, but is "self-taught' and to me, we do not leverage it smartly, or fully. So, I am taking time to bullet a few things we do
that I wanted to hear your thoughts on and also see if anyone here recommends a good "hands-on" type education alternative that we can attend to learn more about how to leverage replication better, and ways to correct some issues we are facing.
Again, new to here, so please forgive me if I do not provide enough information or if I give too much in some cases. Ask what you like and I will try to explain further. I am posting more of statements of issues, and would love input/thoughts to help me
understand how to explain how it is wrong, or understand that it is not and "I" am wrong.
1 - We originally set up a "replication server" to "offload the 'work' from the production servers because reporting services were creating locks and 'bogging down production'.
MY Answer to that is along the lines of, MS Access and badly written queries cause the locks we are seeing, not "report services". LCM type locks are going to lock data, thus we are still seeing locks. Badly coded sql objects that are distributed
either cross-server or cross databases also lend to issues.
2 - A MAJOR one that perplexes me, is having a development --> test --> PROD environment, but coding objects that can only exist on the replicated server, not on the prod server of that line of databases, creates two actual prod databases, each different,
and is highly inadvisable.
An example is a replication server that houses replicated data from more than one prod server/database line, so they can share data without having to use linked server calls, created an environment that has people coding objects that cannot be promoted to
actual PROD, but need to be promoted directly to the replicated database. IMO, that is not manageable. If someone snaps the database again and overwrites, who is to say it does not blow away the non-replicated database?
3 - We have one database that does not have PKs (don't get me started..)...heh.. but it is a 3rd party thing and we need to replicate it. We use transactional repl, but I used Snapshot and it worked, but the TRN file ballooned till it filled up the
drive.
I figure there has to be a way to avoid this, but have not figured that aspects out yet.. heh
Okay, I wrote a lot, and do not want to make anyone angry or offend anyone. I will watch, read, and learn some things, I hope.
Thank you all very much!
R