In today's scale OLTP system, it is a good way to solve the problem by replicating multiple copies in order to alleviate the single point of reading and writing pressure in the database, especially the read pressure.
The SQL Server transactional replication transactional Replication is the most common form of replication we take. Some of the previous articles always introduce some basic constructs,
Lack of some introduction to the optimization, today I will briefly introduce some of them can be adjusted to optimize the place, (in fact, Microsoft has done a good job, the default is to
To meet most of our situation, hope that through the introduction of this article, we combine the actual situation of their system, do some adjustments may have some unexpected harvest, attention in the production
Use caution in the environment)
Before we do that, let's look at the process of copying things.
We can see from publication to subscriber this time through the logreader agent, Distribution Agent these two agents
Used to transfer data changes from the production library, and ultimately to the subscription repository. And for these two agents can be optimized content, in this article
Let's focus on the LogReader Agent first. In the following article, we will continue to introduce other adjustable places.
Logscanthreshold
ReadBatchSize
ReadBatchThreshold
MaxCmdsInTran
Go home and eat first.
Not to be continued ...
SQL Server Transactional replication tuning (1)