So is there any reason we have the reporter in reporter.php instead
of index.php?
Or is this only for hysterical raisins?
For aesthetic reasons, I'd be in favor of moving reporter.php back to
index.php and putting in a redirect file from reporter.php back to
index.php (or, more specificall
I donno. But I would favor the index.php route as well.
Also FYI I'm working on the submit.php script a bit this morning. I'm
doing 2 things:
- Cleaning up error messages
- Inserting a transaction to solve a race condition that was
triggered last night by an IU submit.
-- Josh
On Aug 30
On Aug 30, 2007, at 9:47 AM, Josh Hursey wrote:
I donno. But I would favor the index.php route as well.
K. I'll make the switch.
Also FYI I'm working on the submit.php script a bit this morning. I'm
doing 2 things:
- Cleaning up error messages
- Inserting a transaction to solve a race c
FYI: I finished cleaning up the PHP submit script. It seems that
there are some associated issues with the stats PHP script. I'm
cleaning that up at the moment.
-- Josh
On Aug 30, 2007, at 9:47 AM, Josh Hursey wrote:
I donno. But I would favor the index.php route as well.
Also FYI I'm wor
K. I made the switch index.php <--> reporter.php.
On Aug 30, 2007, at 10:33 AM, Josh Hursey wrote:
FYI: I finished cleaning up the PHP submit script. It seems that
there are some associated issues with the stats PHP script. I'm
cleaning that up at the moment.
-- Josh
On Aug 30, 2007, at 9:47
I just finished cleaning up the warnings in the stats php script.
There is one warning I don't know how to fix:
-
PHP Warning: Missing argument 1 for print_ga() in /nfs/rontok/xraid/
data/osl/www/www.open-mpi.org/mtt/google-analytics.inc on line 6,
referer: http://www.open-mp
Ya, that's me -- I'll fix. Thanks!
On Aug 30, 2007, at 10:52 AM, Josh Hursey wrote:
I just finished cleaning up the warnings in the stats php script.
There is one warning I don't know how to fix:
-
PHP Warning: Missing argument 1 for print_ga() in /nfs/rontok/xraid/
data/osl/w
Ethan --
Did Pak really mean to submit to the database?
On Aug 30, 2007, at 1:22 PM, jjhur...@osl.iu.edu wrote:
Author: jjhursey
Date: 2007-08-30 13:22:06 EDT (Thu, 30 Aug 2007)
New Revision: 964
URL: https://svn.open-mpi.org/trac/mtt/changeset/964
Log:
Increase the 'mpi_version' char length
Pak, we're scratching our heads on this new "developer
results into central database" use case. Could you give more
details?
-Ethan
On Thu, Aug/30/2007 01:32:26PM, Jeff Squyres wrote:
> Ethan --
>
> Did Pak really mean to submit to the database?
>
> On Aug 30, 2007, at 1:22 PM, jjhur...@osl.iu
Oh you mean why would anyone want to submit their intel test results to
the database??
So the scenario is like this, I have a dev ws setup for a pending CMR
for my PML changes. The changes are also in my tmp branch. The point of
running the test to make sure my new dev stuff works before putti
I don't have nightly summary mails from last night -- did anyone else
get them?
--
Jeff Squyres
Cisco Systems
Come to think of it, I didn't get mails this morning, either.
And -- arrgh -- we'll have to fix the index.php <--> reporter.php,
thing, too. But that change was made after 9am this morning, so it
should not have affected last night's / this morning's e-mails.
On Aug 30, 2007, at 3:26 PM,
Shouldn't UH now show up in the summary stats?
http://www.open-mpi.org/mtt/stats/
I only see cisco, iu, sun, ibm, hlrs for a query of 2007-08-01 -
2007-08-30.
--
Jeff Squyres
Cisco Systems
I haven't updated them in a couple days (since monday actually). I
was working on updateing the script so it can run once a day. Should
be done later tonight sometime.
So that's why UH isn't in there yet.
-- josh
On Aug 30, 2007, at 3:35 PM, Jeff Squyres wrote:
Shouldn't UH now show up in
Humm. I think Ethan has this in his crontab and should be able to
check if it is giving errors.
I'm not currently subscribed to the nightly emails, I probably should
do that again.
-- Josh
On Aug 30, 2007, at 3:30 PM, Jeff Squyres wrote:
Come to think of it, I didn't get mails this morni
Oh, I see -- it's a hard-coded list of org's. I thought it got them
from the db.
No rush -- I just saw it today and was curious.
On Aug 30, 2007, at 4:32 PM, Josh Hursey wrote:
I haven't updated them in a couple days (since monday actually). I
was working on updateing the script so it can
cron/alerts.php said this last night:
Sorry, this page is not mirrored. Please see the http://www.open-mpi.org/";>original version of this
page on the main Open MPI web site.
Is "curl" confused by deny_mirror()'s new location (see
r949)?
-Ethan
On Thu, Aug/30/2007 04:34:34PM, Josh Hursey
It could be that we're more stringent now.
What URL are you requesting, specifically?
On Aug 30, 2007, at 4:53 PM, Ethan Mallove wrote:
cron/alerts.php said this last night:
Sorry, this page is not mirrored. Please see the http://www.open-mpi.org/";>original version of this
page on the
Nothing is hardcoded in the collection script, but the orgs are
hardcoded in the stats table once the data has been gathered. That's
why the stats page is so quick to load, since it is referencing a
single relatively small table.
On the stats php page the list of orgs is determined by the d
Josh --
How hard would it be to go change all the existing bakeoff
descriptions in the database to be "[testbake]" in order to
distinguish them from real performance results that we will be
gathering soon "[2007 collective performance bakeoff]"?
--
Jeff Squyres
Cisco Systems
Ethan --
Ok, I think I fixed it the supposed-to-fail problem. I tried for a
while to make a funclet wrapper around perl's map function, but the
syntax for how you had to call it got uglier and uglier. So I gave
up. Instead, I made a &prepend() funclet, and use it like this:
simple_fail
Pretty trivial to do. Just let me know when you want me to do it, and
I can run the update.
On Aug 30, 2007, at 7:19 PM, Jeff Squyres wrote:
Josh --
How hard would it be to go change all the existing bakeoff
descriptions in the database to be "[testbake]" in order to
distinguish them from re
Go ahead and do it now. The idea is that someday soon we'll start
gathering real results and want to be able to distinguish between all
the trials and the real results.
On Aug 30, 2007, at 8:51 PM, Josh Hursey wrote:
Pretty trivial to do. Just let me know when you want me to do it, and
I c
Should be done now. Let me know if you see any residuals floating
around.
-- Josh
On Aug 30, 2007, at 8:54 PM, Jeff Squyres wrote:
Go ahead and do it now. The idea is that someday soon we'll start
gathering real results and want to be able to distinguish between all
the trials and the real
24 matches
Mail list logo
