-
Task
-
Resolution: Won't Do
-
L3 - Default
-
None
-
None
-
Not defined
Context:
Based on the feedback from this change https://github.com/camunda/camunda-optimize/pull/1828#discussion_r396445386 we could think of improving the import efficiency in terms of same timestamp entity handling even further.
Given the following scenario:
Entities: [A, B, C, D1, D2, D3, D4]
Timestamps: [ta, tb, tc, td , td ,td , td]
PageSize: 3
#0:
count = 0
t = 1970 → next: [A, B, C] last: []
#1:
count=1
t = tc → next [D1, D2, D3] last: [C]
#2:
count=3
t = td → next[], last: [D1, D2, D3, D4]
#3:
count=4
t = td → next[], last: [D1, D2, D3, D4]
Current behavior that could be improved is :
round #2 imports D1, D2, and D3 again
Potential Optimization measure: we could keep a Set of the last executions entity id's and exclude already seen ones
round #3 (and all subsequent rounds without data) will fetch last: [D1, D2, D3, D4] again.
Potential Optimization measure: if entities of that same timestamp have already been fetched there is no need to re-fetch them at all in consecutive executions
AT:
- the two optimization cases described are addressed
This is the controller panel for Smart Panels app
- is related to
-
OPT-3456 Do not reimport identical entitiesLastTimestamp batches
-
- Done
-