Have you modified your Service Manager group calc settings as is shown in this post (GroupCalcPollingIntervalMilliseconds)? I’ve found quite a few implementations have been configured with this because it promises better performance in Service Manager. As much as it might be true that this reduces the load of group calculation, it has also proven to be problematic in various implementations. The configuration of group calculation to a very large number can cause problems with SLOs and group permissions allocation. The most significant is the SLOs, which are impacted because the group calculation is involved in the application of work queues to the SLO.
If you are experiencing slow SLO application performance there are two potential options. The first option would be to try reducing the group calculation interval to default and see if it fixes the problem. Another option is to offload the SLO application process, which is more complicated and only used in certain situations.
I hope this helps you out!