[Babase] Re: Upmpi ready-ish

Karl O. Pinc kop at meme.com
Fri Dec 11 15:10:14 EST 2009


On 12/11/2009 01:18:27 PM, Lacey Maryott Roerish wrote:
> Hmm can't delete from MPI_events because data lives in consorts :-\ 
> is
> consorts ONLY for MPI stuff?

Yes.

> 
> On Fri, Dec 11, 2009 at 1:56 PM, Lacey Maryott Roerish
> <lroerish4 at gmail.com>wrote:
> 
> > Hey Karl,
> >
> >    I tried just a small chunk, MID 1- 199 and got quite a few
> errors. It
> > doesn't seem to like 2 Actors on the first line of the interaction. 
> I
> > checked the docs to make sure I was following the format and it
> seems right,
> > it is just saying there are more than 3 characters.
> >
> > Let me know when I should try again or what I should try
> > Thanks
> > L
> >
> >
> > On Fri, Dec 11, 2009 at 1:10 AM, Karl O. Pinc <kop at meme.com> wrote:
> >
> >> Hi Lacey,
> >>
> >> You can try the upmpi program in babase_test.  The
> >> babase database does not have the latest tables etc.
> >>
> >> Attached is the text file I used to create the MPIACTS table rows.
> >> You will also want to create PARTUNKS and CONTEXT_TYPES.
> >>
> >> We may want to talk about unknown individual codes.
> >>
> >> If you get things to upload you may want to VACUUM ANALYZE
> >> before doing any testing.  After the very first upload
> >> (attempt even) things slow down a lot until you do this.
> >> Also don't forget to run queries 2 times and discard the
> >> 1st run because we expect the database to fit in memory
> >> (usually).
> >>
> >> I have not yet added group (via members) into the
> >> MPI_EVENTS view.
> >>
> >> I have also tweaked the upmpi documentation slightly.
> >> Do you have any further comment?  At least check out
> >> the new footnotes.
> >>
> >> Karl <kop at meme.com>
> >> Free Software:  "You don't pay back, you pay forward."
> >>                 -- Robert A. Heinlein
> >>
> >>
> >
> >
> > --
> > - -
> > Lacey K. Maryott Roerish
> > Alberts Lab
> > Department of Biology
> > Duke University
> > ph: 919-660-7306
> > fax: 919-660-7293
> > Lacey.Maryott at duke.edu
> >
> 
> 
> 
> -- 
> - -
> Lacey K. Maryott Roerish
> Alberts Lab
> Department of Biology
> Duke University
> ph: 919-660-7306
> fax: 919-660-7293
> Lacey.Maryott at duke.edu
> 




Karl <kop at meme.com>
Free Software:  "You don't pay back, you pay forward."
                 -- Robert A. Heinlein




More information about the Babase mailing list