May 24, 2010
Participants: Anand, Barseghian, Crawl, Gallagher, Riddle, Staggs, Wang
- james - was able to get crawler caching, stuffing things into postgres. Last crawl I did took 27 hours for 1 site, and then network crashed, and for peter's usecase we'll need to crawl 5. Need to create code to gracefully resume.
- derik - working on 2.0, reporting 2.0 bugs, and reporting and workflow-run-manager documentation.
- sean - workingon engineering view - adding a model type to kepler, ala FSM, for sensor workflows
- jianwu - fixing master-slave bug. there's now a permission policy for each slave, to allow R, Matlab, python, or the external execution actor.
- manish - created working prototype workflow for transfering data from DT to Metacat.
- TODO fix workflow to not put data into production server, ask jing to remove submitted junk data
- TOOD check in workflow to reap svn
- dan - working with DataTurbine
- looking into mirroring - rbnbAdmin is the only tool i've found to do it, and it requires a gui. I've started modifying its code to do it from cmd line. But can't currently get mirroring to work between 2 machines. Will try to mirror on the same machine to see if it's a firewall issue
- looking into DT not keeping data around -- turns out DT only allocates 10 filesets, regardless of archive size, and every detach seals one. Emailed Mark Miller to get a patch from him to fix this