Tom,
my apologies for the wrong paths, I was copying around several accounts,
fixing right now. The other problem I have no idea about, but trying to
find out.
At 10:27 AM 11/1/2001 -0500, Thomas W. Fredian wrote:
>I just started looking into this and just like everything else in this demo
>preparation, the world changed out from under me. Where might I find the
>libraries
>now?
>
>dg0n7$ ls ~keahey/NetLogger-1.6/
>CHANGELOG LICENSE Makefile NLV README README-NLV bin.linux bin.solaris
>build gnuplot scripts snmp src
>
>2 seconds later:
>
>dg0n7$ ls ~keahey/NetLogger-1.6/
>nothing...
>
>Your Makefile in a031:~keahey/publisher refers to
>/nfs/dsl-homes02/keahey/NetLogger-1.6/bin.linux
>but that directory does not exist either.
>
>-tom
>
>
>Kate Keahey wrote:
>
> > Hi,
> >
> > I need help instrumenting the demo with calls to viz app, please read.
> >
> > We are making progress with the application visualizing interactions during
> > the demo to the point where I think it makes sense to instrument the demo.
> > Right now what we mainly see are little red circles appearing in
> > appropriate parts of the country when computation is started, so nothing
> > exciting but it looks like we'll get there. The demo can be instrumented by
> > putting calls reporting on events in the body of demo components when those
> > events happen (so for example when efit writes out data, we report on this
> > event and it will be visualized). There is one problems with instrumenting
> > the demo: unfortunately the Netlogger (which has to be used in order for
> > things to work with this application) does not have a C shell API. I could
> > either write a little program callable from Tom's script or we could push
> > the calls into components for which we do have API. Let me know which you
> > would prefer. I wrote a little program generating events for our demo with
> > lots of comments on how the demo needs to be instrumented (this is the one
> > we use for debugging the app); it is accessible at ~keahey/publish on
> > either the datagrid cluster or sc desksides.
> >
> > Here is how this app works. The interactions are based on a piece of
> > software called Netlogger (see http://www-didc.lbl.gov/NetLogger/). It
> > implements a database where we can post events. Then the vizapp gets the
> > events from there and visualizes them. We need to worry about the first
> > part of the interaction only. Right now and for the next few days the
> > database will be running on terra.mcs.anl.gov reading from port 14835 (see
> > example). At SC it will move to the floor, so it should be configurable.
> >
> > In order to instrument a program, insert calls opening and closing
> > connection to the netlogger at the beginning and end of the program (see
> > example for C API to do this, I will be happy to provide examples in other
> > languages or you can go to the url below). Then insert calls writing events
> > pretty much as they are in the example in the appropriate places (it should
> > be pretty much cut and paste from the example). Compile with netlogger
> > header file and library; on the datagrid cluster and sc desksides the
> > Netlogger installation is in ~keahey/NetLogger-1.6/linux.bin. Mary, I am
> > not sure if you have it on diesel or natasha, installation is essentially
> > just a matter of unpacking. Or I could do it for you if you tell me where.
> >
> > Further info: we are publishing two kinds of events (I took some shortcuts
> > to make things easy so for example bandwidth info is "made up"):
> >
> > 1) JOB_STATUS, format: "ID=%s<not important> FRIENDLY.NAME=%s <demo
> > identifier> JOB.HOST=%s<host where job is running> JOBS.SUBMITTED=%d
> > JOBS.COMPLETE=%d JOBS.FAILED=%d BYTES.PRODUCED=%d"
> >
> > 2) "TransferPerfTotal", format: "ID=%s<not important> FRIENDLY.NAME=%s<demo
> > identifier> URL.SOURCE=%s URL.DEST=%s BYTES=%d BW.CURRENT=%f BW.AVG=%f"
> >
> > Here is the help I need:
> >
> > 1) Tom and Qian, since you know best where the event generating things
> > happen in your programs, could you inert the appropriate calls from the
> > example?
> > 2) Mary, could you handle the installation on diesel and natasha?
> >
> > Kate
> >
> >
> ===============================================================================
> >
> > This message was sent to the SciDAC National Fusion Collaboratory (NFC)
> > workers list nfc-sc01. Visit the Collaboratory at
> > <http://www.fusiongrid.org/>.
> >
> > To unsubscribe from this list, please send a message to
> > majordomo@fusion.gat.com with the following text in the *body* of the
> > message: unsubscribe nfc-sc01
> >
> > David P. Schissel: <schissel@fusion.gat.com>
> <http://fusion.gat.com/~schissel/>
__________________________
Dr. Kate Keahey
Math & Computer Science Div.
Argonne National Laboratory
Argonne, IL 60439, USA
(630) 252-1673
===============================================================================
This message was sent to the SciDAC National Fusion Collaboratory (NFC)
workers list nfc-sc01. Visit the Collaboratory at
<http://www.fusiongrid.org/>.
To unsubscribe from this list, please send a message to
majordomo@fusion.gat.com with the following text in the *body* of the
message: unsubscribe nfc-sc01
David P. Schissel: <schissel@fusion.gat.com> <http://fusion.gat.com/~schissel/>
This archive was generated by hypermail 2.1.1 : Fri Feb 08 2002 - 13:34:29 PST