Difference between revisions of "Staff Priorities - Fabio"
From SCECpedia
Jump to navigationJump to search| Line 2: | Line 2: | ||
* Fling Study | * Fling Study | ||
| − | |||
** We are now 40/42 | ** We are now 40/42 | ||
** Remaining scenarios all running on GEOSYS | ** Remaining scenarios all running on GEOSYS | ||
| Line 9: | Line 8: | ||
** Comparing trunk and 11.2.3 on broadband.usc.edu | ** Comparing trunk and 11.2.3 on broadband.usc.edu | ||
** LP and LA passed, waiting for NR to finish running | ** LP and LA passed, waiting for NR to finish running | ||
| − | |||
| − | |||
* Continued work on adapting run_bbp_parallel scripts to run Fling scenarios in 1 pass | * Continued work on adapting run_bbp_parallel scripts to run Fling scenarios in 1 pass | ||
| − | ** | + | ** Problems running on /scratch |
| − | ** | + | ** Changed to /tmp and was able to run Fling scenario 101 in 46 min (versus 1h34m using regular scripts) |
| − | ** | + | ** Compared new Fling scenario 101 with previous HPCC run (from 11.2.3) and it matches 100% |
| + | ** Running Fling Scenario 104 using new script (previous runtime was 2.5 days) | ||
== Today: == | == Today: == | ||
| Line 20: | Line 18: | ||
* Fling Study | * Fling Study | ||
** Monitoring running scenarios | ** Monitoring running scenarios | ||
| − | * | + | * Package new running scripts and check them into svn as part of BBP |
* Continue working on the testing framework | * Continue working on the testing framework | ||
* Work on Cruise Control setup (pending John Yu install) | * Work on Cruise Control setup (pending John Yu install) | ||
Revision as of 16:22, 27 June 2012
Contents
Yesterday:
- Fling Study
- We are now 40/42
- Remaining scenarios all running on GEOSYS
- Updated wiki tables: http://scec.usc.edu/scecpedia/Fling_Study#Codebase_URS.2FURS.2FURS.2FURS
- Running full validation test for URS
- Comparing trunk and 11.2.3 on broadband.usc.edu
- LP and LA passed, waiting for NR to finish running
- Continued work on adapting run_bbp_parallel scripts to run Fling scenarios in 1 pass
- Problems running on /scratch
- Changed to /tmp and was able to run Fling scenario 101 in 46 min (versus 1h34m using regular scripts)
- Compared new Fling scenario 101 with previous HPCC run (from 11.2.3) and it matches 100%
- Running Fling Scenario 104 using new script (previous runtime was 2.5 days)
Today:
- Fling Study
- Monitoring running scenarios
- Package new running scripts and check them into svn as part of BBP
- Continue working on the testing framework
- Work on Cruise Control setup (pending John Yu install)
Blocked:
- John Yu: Installing Cruise Control
Followups:
- Meeting with Phil to talk about Broadband development priorities @ TBD
Ongoing:
- Look for ways to break up the GF directory layout, some ideas are:
- Figure out how to separate the validation data from Greens Functions
- Create separate validation package
- Separate events and codebases from each other
