From gallen at cct.lsu.edu Mon Oct 5 07:47:01 2009 From: gallen at cct.lsu.edu (Allen Gabrielle) Date: Mon, 5 Oct 2009 07:47:01 -0500 Subject: [Users] XiRel Talk: Tuesday October 6th Message-ID: <1A9A2B78-5E86-40BD-ACD0-54056E671E01@cct.lsu.edu> The next talk in our XiRel series of computational relativity talks will be on October 6th at 10am (Central US time) This week Roland Haas from Georgia Tech will be talking on "Simulating Mixed Binary Systems using Whisky and the BSSN Formalism." The talk will include describing changes to Whisky, sample results, and remaining issues. For information on how to dial in, and to download the presentation just before the talk, please see the web page at http://www.cct.lsu.edu/xirel/Community.html Please forward to anyone who might be interested to join in. Gabrielle -- Gabrielle Allen Associate Professor Department of Computer Science Louisiana State University gallen at cct.lsu.edu -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.cactuscode.org/pipermail/users/attachments/20091005/9d892d22/attachment.html From roland.haas at physics.gatech.edu Thu Oct 15 17:33:59 2009 From: roland.haas at physics.gatech.edu (Roland Haas) Date: Thu, 15 Oct 2009 18:33:59 -0400 Subject: [Users] Cactus/Carpet using mvapich2 on Ranger Message-ID: <4AD7A357.6020202@mail.gatech.edu> Hello all, I just ran into a problem running Cactus on Ranger. For a test I compiled Cactus (using the XiRel codebase, ie. Cactus+Carpet+Whisky) using mvapich2 version 1.2 and the configuration file ranger-mvapich2 from simfactory (revision 275, 2009-02-04; so not the very newest one). When I submit this executable on say 32 cores the code hangs (I believe in MPI_Init) while using up 100% cpu time. Has anybody seen this before and would know how to make it work? Or is mvapich2 1.2 simply not compatible with Cactus? Obviously I can simply use a different mvapich version on Ranger and run with it, so I am just curious, really. Yours, Roland -- My email is as private as my paper mail. I therefore support encrypting and signing email messages. Get my PGP key from . -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: OpenPGP digital signature Url : http://www.cactuscode.org/pipermail/users/attachments/20091015/15ec41ec/attachment.bin From schnetter at cct.lsu.edu Thu Oct 15 18:21:59 2009 From: schnetter at cct.lsu.edu (Erik Schnetter) Date: Thu, 15 Oct 2009 18:21:59 -0500 Subject: [Users] Cactus/Carpet using mvapich2 on Ranger In-Reply-To: <4AD7A357.6020202@mail.gatech.edu> References: <4AD7A357.6020202@mail.gatech.edu> Message-ID: On Oct 15, 2009, at 17:33 , Roland Haas wrote: > Hello all, > > I just ran into a problem running Cactus on Ranger. For a test I > compiled Cactus (using the XiRel codebase, ie. Cactus+Carpet+Whisky) > using mvapich2 version 1.2 and the configuration file ranger- > mvapich2 from simfactory (revision 275, 2009-02-04; so not the very > newest one). > > When I submit this executable on say 32 cores the code hangs (I > believe in MPI_Init) while using up 100% cpu time. > > Has anybody seen this before and would know how to make it work? Or > is mvapich2 1.2 simply not compatible with Cactus? Obviously I can > simply use a different mvapich version on Ranger and run with it, so > I am just curious, really. Roland The configuration options in the simulation factory () should be current, and to the best of my knowledge they are working. Of course, the build options for the executable and the script file used to submit the job must be compatible (using the same MPI version, same compiler libraries etc.) Yes, these options use mvapich2 1.2, together with Intel 10.1. You can use ldd to find out whether the executable loads the right shared libraries. -erik -- Erik Schnetter http://www.cct.lsu.edu/~eschnett/ From knarf at cct.lsu.edu Thu Oct 15 19:36:39 2009 From: knarf at cct.lsu.edu (Frank Loeffler) Date: Thu, 15 Oct 2009 19:36:39 -0500 Subject: [Users] Cactus/Carpet using mvapich2 on Ranger In-Reply-To: <4AD7A357.6020202@mail.gatech.edu> References: <4AD7A357.6020202@mail.gatech.edu> Message-ID: <20091016003639.GU1603@numrel07.cct.lsu.edu> On Thu, Oct 15, 2009 at 06:33:59PM -0400, Roland Haas wrote: > When I submit this executable on say 32 cores the code hangs (I > believe in MPI_Init) while using up 100% cpu time. Hi, this sounds very familiar to some of my current problems. A ticket is submitted and (hopefully) worked on (they asked for binaries, executables, parameter files ect to reproduce it). The whole problem is really odd. The same binary and very similar parameter files seem to work for 1-16 cores, hang for 32 and 64 cores, but again work for >=128 cores. Frank -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 197 bytes Desc: not available Url : http://www.cactuscode.org/pipermail/users/attachments/20091015/372cece2/attachment.bin From roland.haas at physics.gatech.edu Thu Oct 15 19:43:29 2009 From: roland.haas at physics.gatech.edu (Roland Haas) Date: Thu, 15 Oct 2009 20:43:29 -0400 Subject: [Users] Cactus/Carpet using mvapich2 on Ranger In-Reply-To: <20091016003639.GU1603@numrel07.cct.lsu.edu> References: <4AD7A357.6020202@mail.gatech.edu> <20091016003639.GU1603@numrel07.cct.lsu.edu> Message-ID: <4AD7C1B1.30505@mail.gatech.edu> Hello Frank, > this sounds very familiar to some of my current problems. A ticket is > submitted and (hopefully) worked on (they asked for binaries, > executables, parameter files ect to reproduce it). > The whole problem is really odd. The same binary and very similar > parameter files seem to work for 1-16 cores, hang for 32 and 64 cores, > but again work for >=128 cores. Yes, this is precisely what I see (at least the 1-16,32,64 part, I have not tried 128 cores). This happens not just on ranger but also on a different mvapich2 1.2 installation [used to benchmark a cluster]. Thank you for letting me no. Yours, Roland -- My email is as private as my paper mail. I therefore support encrypting and signing email messages. Get my PGP key from . -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: OpenPGP digital signature Url : http://www.cactuscode.org/pipermail/users/attachments/20091015/e5927bc8/attachment.bin From barry.wardell at aei.mpg.de Wed Oct 28 07:39:23 2009 From: barry.wardell at aei.mpg.de (Barry Wardell) Date: Wed, 28 Oct 2009 14:39:23 +0100 Subject: [Users] Write.c missing from IOJpeg thorn? Message-ID: <4AE8498B.606@aei.mpg.de> Hi, I have been trying to compile the IOJpeg thorn. When I do so I get an error about not being able to find Write.c: No rule to make target `Write.c.o', needed by `make.checked'. Stop. The make.code.defn file refers to Write.c, but there is no such file in the IOJpeg/src directory I just checked out from cactusdevcvs. Does anyone know where the Write.c file has gone? Barry Wardell From schnetter at cct.lsu.edu Wed Oct 28 07:46:34 2009 From: schnetter at cct.lsu.edu (Erik Schnetter) Date: Wed, 28 Oct 2009 08:46:34 -0500 Subject: [Users] Write.c missing from IOJpeg thorn? In-Reply-To: <4AE8498B.606@aei.mpg.de> References: <4AE8498B.606@aei.mpg.de> Message-ID: <6D23EBC5-D56F-4BAF-A5C2-4393F1E1E845@cct.lsu.edu> On Oct 28, 2009, at 8:39 , Barry Wardell wrote: > Hi, > > I have been trying to compile the IOJpeg thorn. When I do so I get an > error about not being able to find Write.c: > > No rule to make target `Write.c.o', needed by `make.checked'. > Stop. > > The make.code.defn file refers to Write.c, but there is no such file > in > the IOJpeg/src directory I just checked out from cactusdevcvs. Does > anyone know where the Write.c file has gone? Barry The file should be there. It should have version 1.19 in cvs. I suspect there may have been a problem with your checkout, or maybe a problem with your file system. -erik -- Erik Schnetter http://www.cct.lsu.edu/~eschnett/ From knarf at cct.lsu.edu Wed Oct 28 07:48:34 2009 From: knarf at cct.lsu.edu (Frank Loeffler) Date: Wed, 28 Oct 2009 08:48:34 -0500 Subject: [Users] Write.c missing from IOJpeg thorn? In-Reply-To: <6D23EBC5-D56F-4BAF-A5C2-4393F1E1E845@cct.lsu.edu> References: <4AE8498B.606@aei.mpg.de> <6D23EBC5-D56F-4BAF-A5C2-4393F1E1E845@cct.lsu.edu> Message-ID: <20091028134833.GI22099@numrel07.cct.lsu.edu> On Wed, Oct 28, 2009 at 08:46:34AM -0500, Erik Schnetter wrote: > The file should be there. It should have version 1.19 in cvs. I > suspect there may have been a problem with your checkout, or maybe a > problem with your file system. I just updated my checkout and the file really went missing. It also does not seem to have a 'history' in CVS anymore. I cannot find a commit message deleting this file. I will contact Colby to have him have a look at the repository itself. Frank -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 197 bytes Desc: not available Url : http://www.cactuscode.org/pipermail/users/attachments/20091028/8a6a08c9/attachment.bin From knarf at cct.lsu.edu Wed Oct 28 19:20:59 2009 From: knarf at cct.lsu.edu (Frank Loeffler) Date: Wed, 28 Oct 2009 20:20:59 -0500 Subject: [Users] Write.c missing from IOJpeg thorn? In-Reply-To: <4AE8498B.606@aei.mpg.de> References: <4AE8498B.606@aei.mpg.de> Message-ID: <20091029012059.GV22099@numrel07.cct.lsu.edu> Hi, On Wed, Oct 28, 2009 at 02:39:23PM +0100, Barry Wardell wrote: > I have been trying to compile the IOJpeg thorn. When I do so I get an > error about not being able to find Write.c: This really seems to be a problem with the repository. While we are investigating you can find the most recent (?) version of the file here: http://www.cct.lsu.edu/~knarf/Write.c We are sorry for the inconvenience, Frank -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 197 bytes Desc: not available Url : http://www.cactuscode.org/pipermail/users/attachments/20091028/e6a045e6/attachment.bin From barry.wardell at aei.mpg.de Thu Oct 29 04:36:03 2009 From: barry.wardell at aei.mpg.de (Barry Wardell) Date: Thu, 29 Oct 2009 11:36:03 +0100 Subject: [Users] Write.c missing from IOJpeg thorn? In-Reply-To: <20091029012059.GV22099@numrel07.cct.lsu.edu> References: <4AE8498B.606@aei.mpg.de> <20091029012059.GV22099@numrel07.cct.lsu.edu> Message-ID: <4AE97013.2090503@aei.mpg.de> Frank Loeffler wrote: > This really seems to be a problem with the repository. While we are > investigating you can find the most recent (?) version of the file here: > > http://www.cct.lsu.edu/~knarf/Write.c > > We are sorry for the inconvenience, Frank > Thanks, using this file I am now able to compile everything I need. A possibly related issue I have encountered is with the stable repository. When I checked out the flesh, the file src/include/cctk.h was missing with no CVS history information. I think there were also some other files missing from src/include/. Is there also a problem with this repository? Barry From bentivegna at cct.lsu.edu Thu Oct 29 17:11:59 2009 From: bentivegna at cct.lsu.edu (Eloisa Bentivegna) Date: Thu, 29 Oct 2009 18:11:59 -0500 Subject: [Users] [Fwd: [Allcct] Cactus CVS Maintenance] Message-ID: <4AEA213F.4070805@cct.lsu.edu> -------- Messaggio Originale -------- Oggetto: [Allcct] Cactus CVS Maintenance Data: Thu, 29 Oct 2009 18:00:46 -0500 Da: Adam Yates A: allcct at cct.lsu.edu Hi All; We are performing some maintenance on the CVS Cactus repository for the next hour (6-7pm CST). Please refrain from checking files in or out during this time period so we can make sure we don't miss any files. Another email will follow when the maintenance has been completed. Thanks for your patience! -- Adam Yates Systems Administrator -- Research Infrastructure Center for Computation and Technology 232 Johnston Hall, Baton Rouge, LA 70803 W: 225.578.8235 C: 225.663.0218 From bentivegna at cct.lsu.edu Thu Oct 29 20:54:51 2009 From: bentivegna at cct.lsu.edu (Eloisa Bentivegna) Date: Thu, 29 Oct 2009 21:54:51 -0500 Subject: [Users] [Fwd: Re: [Allcct] Cactus CVS Maintenance] Message-ID: <4AEA557B.4030008@cct.lsu.edu> -------- Messaggio Originale -------- Oggetto: Re: [Allcct] Cactus CVS Maintenance Data: Thu, 29 Oct 2009 18:55:18 -0500 Da: Adam Yates A: allcct at cct.lsu.edu Referenze: <4AEA1E9E.9030607 at cct.lsu.edu> Maintenance has been completed on the Cactus CVS repository. Please let us know if you have any more troubles. Thanks! Adam -- Adam Yates Systems Administrator -- Research Infrastructure Center for Computation and Technology 232 Johnston Hall, Baton Rouge, LA 70803 W: 225.578.8235 C: 225.663.0218 From barry.wardell at aei.mpg.de Fri Oct 30 07:35:15 2009 From: barry.wardell at aei.mpg.de (Barry Wardell) Date: Fri, 30 Oct 2009 14:35:15 +0100 Subject: [Users] [Developers] [Fwd: Re: [Allcct] Cactus CVS Maintenance] In-Reply-To: <4AEA557B.4030008@cct.lsu.edu> References: <4AEA557B.4030008@cct.lsu.edu> Message-ID: <606BCEA8-5743-4A12-8814-97E835A7F2FD@aei.mpg.de> On 30 Oct 2009, at 03:54, Eloisa Bentivegna wrote: > Maintenance has been completed on the Cactus CVS repository. > > Please let us know if you have any more troubles. Was this maintenance in response to my report of a missing file (arrangements/CactusIO/IOJpeg/Write.c) from the repository? I was hoping the file would be restored after the maintenance was completed, but I've just done a fresh checkout and it is still missing. Barry From bentivegna at cct.lsu.edu Fri Oct 30 10:15:54 2009 From: bentivegna at cct.lsu.edu (Eloisa Bentivegna) Date: Fri, 30 Oct 2009 11:15:54 -0500 Subject: [Users] [Developers] [Fwd: Re: [Allcct] Cactus CVS Maintenance] In-Reply-To: <606BCEA8-5743-4A12-8814-97E835A7F2FD@aei.mpg.de> References: <4AEA557B.4030008@cct.lsu.edu> <606BCEA8-5743-4A12-8814-97E835A7F2FD@aei.mpg.de> Message-ID: <4AEB113A.4010801@cct.lsu.edu> Barry Wardell ha scritto: > Was this maintenance in response to my report of a missing file > (arrangements/CactusIO/IOJpeg/Write.c) from the repository? I was > hoping the file would be restored after the maintenance was completed, > but I've just done a fresh checkout and it is still missing. Yes. It turns out that the maintainance was not in fact complete; they are currently working on this. Thanks for reporting the issue! Eloisa