| Time | Item | Who | Notes |
|---|
| ~20min | Tier 1 Cadence Runs | Cathy | - Kem has been executing runs on both opsimblitz1 and opsimblitz2 which completed at the end of last weekend. Some messy work-arounds were necessary as users kcook and lsst are not able to run main.py on opsimblitz2 (and likewise opsimblitz1 is restricted to only kcook). A JIRA item has been created (OPSIM-145) and will be addressed by holding a special meeting Mar 5 to discuss the range of multi-user environment issues we are experiencing. These will need to be settled before we can deploy MAF.
- Also, the summaryreport.py script on opsimblitz1 is taking 17.5 hours by itself to complete and it is not clear why. The work-around is to use the opsimblitz2 runs where the summaryreport.py take just under 2 hours, and the entire post-processing script takes about 6.5 hours. Since running more than one SSTAR at a time does not complete any SSTAR (see set of 11 SSTARs running on opsimblitz1 for the past 6 days), these SSTARs are set to run sequentially and the Tier 1 set of runs should be ready for distribution on Friday.
- Cathy will post the links to the opsimcvs machine on a Confluence page, and Peter will let us know if other more convenient formats or data bundles are needed.
|
| ~10min | Workshop Dry Run Jun 9,10 | Zeljko | - Motivation is to ensure the quality and completeness of the presentations for the Cadence Workshop, to create presentations with a hands-on component and test them out, and to include the sci collabs members who will be helping with the hands-on part in developing the material.
- Zeljko, Lynne & Peter plan to come to Tucson with the Sci Collabs volunteers (to be invited) to work on this for several days)
- Zeljko is leading the Dry Run planning; Knut and Steve are preparing a document which defines the expectations, agenda and deliverables for the Workshop which will be approved by both NOAO and the LSST project.
- OpSim will deliver the Tier 1 runs this week, and it is possible that some variations of runs will be requested as followup analyses are being completed.
|
| ~5min | Scheduler Planning | Abi | - Post-Review, we are shifting from the paradigm of building all the required capability into v3.0 with hooks to the OCS, then rewriting the necessary algorithms into a final Scheduler in an appropriate language, to a new paradigm of continuing development of the simulator code with an eye towards evolving it to a codebase with an extractable Blue Box scheduler, and in parallel building an OCS emulator which will accept the Blue Box.
- A planning document detailing the components and considerations for delivering a Scheduler is under development.
|
| ~5min | MAF status | Peter | - Alpha version should be released by the end of February
- Will include SSTAR replica - Cathy will detail additions and improvements planned for SSTAR on a Confluence page for Peter & Lynne to include as upgrades to the existing SSTAR sections.
|
