Sizing guidelines
Depending on the number of Replicate tasks and concurrent users, you might need to balance the system load between multiple Enterprise Manager machines. For example, if your hardware configuration is set up for a large system (40 Replicate servers), it is recommended to monitor no more than 4000 Replicate tasks on a single Enterprise Manager machine. When you near this threshold, then the recommendation is to split the tasks between at least two Enterprise Manager machines by dividing the number of monitored Replicate servers between the Enterprise Manager machines. Similarly, if your hardware configuration is set up for a medium system and the number of Replicate tasks and users concurrently accessing the system is approaching the maximum threshold (see table below), then best practice is to split the monitored Replicate servers between at least two Enterprise Manager machines.
For information on hardware configuration guidelines, see Hardware configuration guidelines.
Medium System | Large System | |
---|---|---|
Replicate servers |
8 |
40 |
Tasks |
800 |
4000 |
Concurrent users (Rate of public API calls may affect these numbers) |
10 |
80 |