| | 31 | The bulk of my work this week was on code to speed-up the ingestion of DVO data into MySQL while not breaking the existing mechanism of using the C-API for reading DVO data on an exposure-by-exposure basis, as this will still be more efficient for parts of the sky with thin coverage. One big speed-up came from removing the need for 'SOURCE_ID', which I had been carrying unnecessarily per detection in my MySQL database. |
| | 32 | |
| | 33 | The other development task was figuring out how to queue available data to load to PSPS by region on the sky based on the ratio of exposures I have to load and the size on disk of the corresponding DVO FITS files. An algorithm is required to decide when it is appropriate to load on-the-fly, exposure-by-exposure, and when it is more efficient to pre-ingest a region of DVO. This becomes especially confusing as I have a multi-client system where clients, thus far, worked independently of one another. This will have to change. |
| | 34 | |
| | 35 | Other work included: |
| | 36 | |
| | 37 | * after a couple of weeks of testing new features, I released new version of PSVO |
| | 38 | * tested new STILTS release with ippToPsps |
| | 39 | |
