Hi, our Admin have noticed a spike on CPU usage when replication occurs. They have suggested that we should set the Discard index to “after each use” on views as this will prevent the index from being rebuilt. For commonly used views I’m not so sold on the idea as the user will have to wait as the view index has to be built before it opens. I’m also assuming that this request would still need to be sent to the nupdate.exe queue which still means CPU usage. I do understand that setting the discard to “after each use” will save disk space but does it also help to reduce or stagger CPU usage or spikes.Your thoughts are most welcome.
Thx.