-
Type: Improvement
-
Resolution: Won't Do
-
Priority: Major - P3
-
None
-
Affects Version/s: None
-
Component/s: None
-
None
-
Atlas Streams
Today we send each operator provided state record to the FileIO thread (it is a move of a BSONObj). If we are dealing with a large number of small BSONObjs then this can become a bottleneck. We are expecting that this will not be the case - i.e. operators will do some internal batching on their own.
If this assumption turns out to not be valid in practice, then we may have to revisit this issue.