SQL Server Performance Forum – Threads Archive

replication max number of jobs

We currently have one SQL Server that is being used as the central SQL Distribution agent for all of our replication jobs. However, we have long since passed the Microsoft recommended maximum number of SQL replication jobs on this server. We’ve reconfigured the memory alloted to the SQL server as it was starting to not have enough memory from the OS for it to have any more jobs added. I was curious what is the industry’s "best standard practice" for handling issues like this. I would think that there has to be something better then building a new distribution SQL server every time you approach 100 jobs. Does anybody have any suggestions on how you’ve handled having a large number of replication jobs on a single SQL server. Thanks.

I don’t see there is any POINT to set as best practices for scheduling the jobs including the replication ones. The more jobs you have (even for replication) on a server and the more times they run means the more overall lines that are written to the log. There is a option there for each jobs retention but as they all add up and reach the max for the log itself they oldest items get pushed out first. This is why you would see a difference even if the settings are the same. So you must check the server settings such as Memory and any physical disk issues in order to execute the jobs perfectly and let SQL handle the memory dynamically rather than min and max settings. Satya SKJ
Moderator
http://www.SQL-Server-Performance.Com/forum
This posting is provided â€œAS ISâ€ with no rights for the sake of knowledge sharing.

]]>