Re: [Monetdb-developers] [Monetdb-checkins] testing/src Mtest.py.in, , 1.26, 1.27
On Thu, Jul 16, 2009 at 12:54:20PM +0000, Fabian wrote:
Update of /cvsroot/monetdb/testing/src In directory 23jxhf1.ch3.sourceforge.com:/tmp/cvs-serv5974/src
Modified Files: Mtest.py.in Log Message: Have you ever wanted your machine to nearly die under a load of 20+ by running Mtest? Now you can, using --jobs=<huge-number>.
Implemented Parallel Mtest (r), parallelising where no Mtest has tested before, using X worker jobs, where X defaults to 1.
Parallelisation currently is done over directories only (inter-directory dependencies are hard to track). No more chdirs are performed, so all arguments are made absolute paths, resulting in some long output paths here and there, compared to the old behaviour of Mtest.
This might require quite a bunch of tests to be adapted accordingly, because so far basically all tests assumed that they were executed in the TSTTRGDIR (though there are tests that "don't care")... ... you already fixed some; but I guess there are (many?) more to do ... Wasn't there a chance to do a per-thread chdir? Or does Python not allow that? Stefan
Please test, enjoy and fix (preferably in parallel ;) )
U Mtest.py.in Index: Mtest.py.in =================================================================== RCS file: /cvsroot/monetdb/testing/src/Mtest.py.in,v retrieving revision 1.26 retrieving revision 1.27 diff -u -d -r1.26 -r1.27 --- Mtest.py.in 3 Jun 2009 11:42:33 -0000 1.26 +++ Mtest.py.in 16 Jul 2009 12:54:18 -0000 1.27 @@ -46,6 +46,9 @@ import socket import struct import select +from threading import Thread +import copy +import string #import signal #import getpass import Mfilter @@ -107,18 +110,22 @@ [...1251 lines suppressed...] + w.join() + t = w.getTime() + elem = w.getElem() t_ = t_ + t if elem is not None: body.append(elem) @@ -3756,7 +3832,7 @@ if THISFILE == "Mtest.py": env['TSTDIR'] = "" env['TSTTRGDIR'] = os.path.join(env['TSTTRGBASE'],TSTPREF) - CreateHtmlIndex(env, *body) + CreateHtmlIndex(htmlenv, env, *body)
Failed = 0 for f in Failure[1:]: @@ -3843,3 +3919,4 @@
# END ############################################################################# +# vim: set ts=4 sw=4 expandtab :
------------------------------------------------------------------------------ Enter the BlackBerry Developer Challenge This is your chance to win up to $100,000 in prizes! For a limited time, vendors submitting new applications to BlackBerry App World(TM) will have the opportunity to enter the BlackBerry Developer Challenge. See full prize details at: http://p.sf.net/sfu/Challenge _______________________________________________ Monetdb-checkins mailing list Monetdb-checkins@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/monetdb-checkins
-- | Dr. Stefan Manegold | mailto:Stefan.Manegold@cwi.nl | | CWI, P.O.Box 94079 | http://www.cwi.nl/~manegold/ | | 1090 GB Amsterdam | Tel.: +31 (20) 592-4212 | | The Netherlands | Fax : +31 (20) 592-4312 |
On 17-07-2009 17:30:26 +0200, Stefan Manegold wrote:
This might require quite a bunch of tests to be adapted accordingly, because so far basically all tests assumed that they were executed in the TSTTRGDIR (though there are tests that "don't care")... ... you already fixed some; but I guess there are (many?) more to do ...
Wasn't there a chance to do a per-thread chdir? Or does Python not allow that?
chdir is per process, not thread.
On Fri, Jul 17, 2009 at 05:37:17PM +0200, Fabian Groffen wrote:
On 17-07-2009 17:30:26 +0200, Stefan Manegold wrote:
This might require quite a bunch of tests to be adapted accordingly, because so far basically all tests assumed that they were executed in the TSTTRGDIR (though there are tests that "don't care")... ... you already fixed some; but I guess there are (many?) more to do ...
Wasn't there a chance to do a per-thread chdir? Or does Python not allow that?
chdir is per process, not thread.
What about setting cwd=<TSTTRGDIR> with the respective subprocess.Popen() call(s)? Stefan
------------------------------------------------------------------------------ Enter the BlackBerry Developer Challenge This is your chance to win up to $100,000 in prizes! For a limited time, vendors submitting new applications to BlackBerry App World(TM) will have the opportunity to enter the BlackBerry Developer Challenge. See full prize details at: http://p.sf.net/sfu/Challenge _______________________________________________ Monetdb-developers mailing list Monetdb-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/monetdb-developers
-- | Dr. Stefan Manegold | mailto:Stefan.Manegold@cwi.nl | | CWI, P.O.Box 94079 | http://www.cwi.nl/~manegold/ | | 1090 GB Amsterdam | Tel.: +31 (20) 592-4212 | | The Netherlands | Fax : +31 (20) 592-4312 |
On 17-07-2009 17:57:34 +0200, Stefan Manegold wrote:
On Fri, Jul 17, 2009 at 05:37:17PM +0200, Fabian Groffen wrote:
On 17-07-2009 17:30:26 +0200, Stefan Manegold wrote:
This might require quite a bunch of tests to be adapted accordingly, because so far basically all tests assumed that they were executed in the TSTTRGDIR (though there are tests that "don't care")... ... you already fixed some; but I guess there are (many?) more to do ...
Wasn't there a chance to do a per-thread chdir? Or does Python not allow that?
chdir is per process, not thread.
What about setting cwd=<TSTTRGDIR> with the respective subprocess.Popen() call(s)?
Might be possible, but isn't the current solution more elegant, because its agnostic to the CWD?
participants (2)
-
Fabian Groffen
-
Stefan Manegold