Changes between Version 6 and Version 7 of edelweiss
- Timestamp:
- 05/15/14 08:40:49 (11 years ago)
Legend:
- Unmodified
- Added
- Removed
- Modified
-
edelweiss
v6 v7 20 20 = Note = 21 21 22 === Resample work === 23 22 24 Resample works, but very very slow. Why? Because it is querying each data row per http request, for a aggregation with wide interval, this should not be a problem, but as the interval gets smaller, it will be very slow. 23 25 24 26 Solution: Requires a mapreduce function to do aggregation 27 28 Status: Pending 29 30 === Data Query Speed === 31 32 Currently the speed of caching is roughly 250kB/s on localhost. Pretty slow. Current approach uses a buffer to request 10 thousand data rows per http request. I feel that this can somewhat still be improve. 33 34 Idea: Uses socket 35 36 Idea: use multi exec curl 25 37 26 38 Status: Pending