Does anyone know how to successfully limit the number of SLP replication jobs, specifically when dealing with MSDP? I have read that by limiting the number of I/O streams on the target site, but this does not seem to work.
I have increased the minimum size of the jobs as a way to create fewer of them, but the bottom line is that eventually there are enough replication jobs to stop the regular backup jobs from running. I have noticed that replication jobs will reach a limit on the target, and instead of being queued, they are active against the source MSDP, sucking up an I/O stream, but not moving any data, claiming that the target has reached its limit.
I understand I can disable the SLP's during busy times, but I don't want that either, I just want there to be fewer jobs running. It seems my only viable alternative is to make the minimum job enormous to limit the number, but this doesn't seem like a good solution to me.
Some sort of limit to the number of replication streams should be available in the SLP. As wonderful as SLPs are, they really are lacking a lot of flexibility that would seem pretty straightforward from a design perspective.
Sorry for the rant, but anyone have any other ideas?
DT