From andrescodas at gmail.com Fri Dec 14 21:39:37 2018 From: andrescodas at gmail.com (Andres Codas) Date: Fri, 14 Dec 2018 19:39:37 -0200 Subject: [Opm] Segmentation fault using openporousmedia/opmreleases:latest Message-ID: Hi I get the error bellow when using openporousmedia/opmreleases:latest. The problem runs with openporousmedia/opmreleases:2017.04 Is it a known issue? How can I solve it? Cheers, Andres docker run --rm -t -v "${PWD}":/shared_host openporousmedia/opmreleases:latest flow --output-dir="/shared_host/output/" /shared_host/deck.DATA libibverbs: Warning: couldn't open config directory '/etc/libibverbs.d'. libibverbs: Warning: no userspace device-specific driver found for /sys/class/infiniband_verbs/uverbs0 -------------------------------------------------------------------------- [[56069,1],0]: A high-performance Open MPI point-to-point messaging module was unable to find any relevant network interfaces: Module: OpenFabrics (openib) Host: e3a21ed3293e Another transport will be used instead, although this may result in lower performance. -------------------------------------------------------------------------- ********************************************************************** * * * This is flow 2018.10 * * * * Flow is a simulator for fully implicit three-phase black-oil flow, * * including solvent and polymer capabilities. * * For more information, see https://opm-project.org * * * ********************************************************************** [e3a21ed3293e:00001] *** Process received signal *** [e3a21ed3293e:00001] Signal: Segmentation fault (11) [e3a21ed3293e:00001] Signal code: Address not mapped (1) [e3a21ed3293e:00001] Failing at address: 0x8 [e3a21ed3293e:00001] [ 0] /lib/x86_64-linux-gnu/libpthread.so.0(+0x11390)[0x7f3338197390] [e3a21ed3293e:00001] [ 1] /usr/lib/x86_64-linux-gnu/libecl.so.2(fortio_get_FILE+0x0)[0x7f3336adb800] [e3a21ed3293e:00001] [ 2] /usr/lib/x86_64-linux-gnu/libecl.so.2(_Z20ecl_kw_fwrite_headerPK13ecl_kw_structP13fortio_struct+0x2a)[0x7f3336acf6da] [e3a21ed3293e:00001] [ 3] /usr/lib/x86_64-linux-gnu/libecl.so.2(ecl_kw_fwrite+0x2a)[0x7f3336ad108a] [e3a21ed3293e:00001] [ 4] /usr/lib/x86_64-linux-gnu/libecl.so.2(ecl_init_file_fwrite_header+0x142)[0x7f3336ab9592] [e3a21ed3293e:00001] [ 5] /usr/lib/x86_64-linux-gnu/libopmcommon.so.2018(_ZNK3Opm9EclipseIO4Impl13writeINITFileERKNS_4data8SolutionESt3mapINSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEESt6vectorIiSaIiEESt4lessISC_ESaISt4pairIKSC_SF_EEERKNS_3NNCE+0x1e6)[0x7f3337611726] [e3a21ed3293e:00001] [ 6] /usr/lib/x86_64-linux-gnu/libopmcommon.so.2018(_ZN3Opm9EclipseIO12writeInitialENS_4data8SolutionESt3mapINSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEESt6vectorIiSaIiEESt4lessIS9_ESaISt4pairIKS9_SC_EEERKNS_3NNCE+0x13b)[0x7f333761248b] [e3a21ed3293e:00001] [ 7] flow(_ZN5Ewoms10EclProblemINS_10Properties4TTag22EclFlowOilWaterProblemEE10finishInitEv+0x93e)[0xbc230e] [e3a21ed3293e:00001] [ 8] flow(_ZN5Ewoms9SimulatorINS_10Properties4TTag22EclFlowOilWaterProblemEEC1Eb+0x1946)[0xbc5a36] [e3a21ed3293e:00001] [ 9] flow(_ZN3Opm12FlowMainEbosIN5Ewoms10Properties4TTag22EclFlowOilWaterProblemEE18setupEbosSimulatorEv+0x3b)[0xbd733b] [e3a21ed3293e:00001] [10] flow(_ZN3Opm12FlowMainEbosIN5Ewoms10Properties4TTag22EclFlowOilWaterProblemEE7executeEiPPc+0xdc)[0xbe8b8c] [e3a21ed3293e:00001] [11] flow(_ZN3Opm20flowEbosOilWaterMainEiPPc+0x130)[0xb1f7c0] [e3a21ed3293e:00001] [12] flow(main+0x805)[0x75c845] [e3a21ed3293e:00001] [13] /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf0)[0x7f3335bdc830] [e3a21ed3293e:00001] [14] flow(_start+0x29)[0x76a5a9] [e3a21ed3293e:00001] *** End of error message *** From andrescodas at gmail.com Mon Dec 17 12:14:08 2018 From: andrescodas at gmail.com (Andres Codas) Date: Mon, 17 Dec 2018 10:14:08 -0200 Subject: [Opm] Segmentation fault using openporousmedia/opmreleases:latest In-Reply-To: References: Message-ID: Hi! Please disregard this issue. Although the warnings continue to appear I'm no longer able to reproduce the segfault. I presume they are not related Regards, Andres Den fre. 14. des. 2018 kl. 19:39 skrev Andres Codas : > Hi > I get the error bellow when using openporousmedia/opmreleases:latest. > The problem runs with openporousmedia/opmreleases:2017.04 > > Is it a known issue? How can I solve it? > Cheers, > Andres > > > docker run --rm -t -v "${PWD}":/shared_host > openporousmedia/opmreleases:latest flow > --output-dir="/shared_host/output/" /shared_host/deck.DATA > libibverbs: Warning: couldn't open config directory '/etc/libibverbs.d'. > libibverbs: Warning: no userspace device-specific driver found for > /sys/class/infiniband_verbs/uverbs0 > -------------------------------------------------------------------------- > [[56069,1],0]: A high-performance Open MPI point-to-point messaging module > was unable to find any relevant network interfaces: > > Module: OpenFabrics (openib) > Host: e3a21ed3293e > > Another transport will be used instead, although this may result in > lower performance. > -------------------------------------------------------------------------- > ********************************************************************** > * * > * This is flow 2018.10 * > * * > * Flow is a simulator for fully implicit three-phase black-oil flow, * > * including solvent and polymer capabilities. * > * For more information, see https://opm-project.org * > * * > ********************************************************************** > > [e3a21ed3293e:00001] *** Process received signal *** > [e3a21ed3293e:00001] Signal: Segmentation fault (11) > [e3a21ed3293e:00001] Signal code: Address not mapped (1) > [e3a21ed3293e:00001] Failing at address: 0x8 > [e3a21ed3293e:00001] [ 0] > /lib/x86_64-linux-gnu/libpthread.so.0(+0x11390)[0x7f3338197390] > [e3a21ed3293e:00001] [ 1] > /usr/lib/x86_64-linux-gnu/libecl.so.2(fortio_get_FILE+0x0)[0x7f3336adb800] > [e3a21ed3293e:00001] [ 2] > /usr/lib/x86_64-linux-gnu/libecl.so.2(_Z20ecl_kw_fwrite_headerPK13ecl_kw_structP13fortio_struct+0x2a)[0x7f3336acf6da] > [e3a21ed3293e:00001] [ 3] > /usr/lib/x86_64-linux-gnu/libecl.so.2(ecl_kw_fwrite+0x2a)[0x7f3336ad108a] > [e3a21ed3293e:00001] [ 4] > /usr/lib/x86_64-linux-gnu/libecl.so.2(ecl_init_file_fwrite_header+0x142)[0x7f3336ab9592] > [e3a21ed3293e:00001] [ 5] > /usr/lib/x86_64-linux-gnu/libopmcommon.so.2018(_ZNK3Opm9EclipseIO4Impl13writeINITFileERKNS_4data8SolutionESt3mapINSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEESt6vectorIiSaIiEESt4lessISC_ESaISt4pairIKSC_SF_EEERKNS_3NNCE+0x1e6)[0x7f3337611726] > [e3a21ed3293e:00001] [ 6] > /usr/lib/x86_64-linux-gnu/libopmcommon.so.2018(_ZN3Opm9EclipseIO12writeInitialENS_4data8SolutionESt3mapINSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEESt6vectorIiSaIiEESt4lessIS9_ESaISt4pairIKS9_SC_EEERKNS_3NNCE+0x13b)[0x7f333761248b] > [e3a21ed3293e:00001] [ 7] > flow(_ZN5Ewoms10EclProblemINS_10Properties4TTag22EclFlowOilWaterProblemEE10finishInitEv+0x93e)[0xbc230e] > [e3a21ed3293e:00001] [ 8] > flow(_ZN5Ewoms9SimulatorINS_10Properties4TTag22EclFlowOilWaterProblemEEC1Eb+0x1946)[0xbc5a36] > [e3a21ed3293e:00001] [ 9] > flow(_ZN3Opm12FlowMainEbosIN5Ewoms10Properties4TTag22EclFlowOilWaterProblemEE18setupEbosSimulatorEv+0x3b)[0xbd733b] > [e3a21ed3293e:00001] [10] > flow(_ZN3Opm12FlowMainEbosIN5Ewoms10Properties4TTag22EclFlowOilWaterProblemEE7executeEiPPc+0xdc)[0xbe8b8c] > [e3a21ed3293e:00001] [11] > flow(_ZN3Opm20flowEbosOilWaterMainEiPPc+0x130)[0xb1f7c0] > [e3a21ed3293e:00001] [12] flow(main+0x805)[0x75c845] > [e3a21ed3293e:00001] [13] > /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf0)[0x7f3335bdc830] > [e3a21ed3293e:00001] [14] flow(_start+0x29)[0x76a5a9] > [e3a21ed3293e:00001] *** End of error message *** > From sindimo at gmail.com Tue Dec 18 21:09:01 2018 From: sindimo at gmail.com (M. S.) Date: Tue, 18 Dec 2018 16:09:01 -0500 Subject: [Opm] Large Models for OPM Flow Message-ID: Dear All, I was wondering what's the largest number of processing cores that Flow has been tested with for the MPI parallel version, also if there's a repository having large models (e.g. 1 million cells or larger) ready to test Flow with for benchmarking on 16, 32, 64,128, etc cores. I am currently using the Norne and SPE9 models for testing but they're too small to be run on larger number of cores. I was also previously told here that the larger SPE10 model 2 may not be the best model to use for benchmarking as it's currently not included in any of the integration testing and not the most stable to run. For example ECLIPSE provides a model ONEM1 with different variations for benchmarking on different number of cores (e.g. ONEM1_8, ONEM1_16, etc...), I tried running those models using Flow but it seems that the model input syntax for ECLIPSE isn't 100% compatible with Flow. Any help on this would be much appreciated. Many thanks. Sincerely, Mohamad MIT Geonumerics group From abir at equinor.com Tue Dec 18 22:12:06 2018 From: abir at equinor.com (Alf Birger Rustad) Date: Tue, 18 Dec 2018 22:12:06 +0000 Subject: [Opm] Large Models for OPM Flow In-Reply-To: References: Message-ID: Hi Mohamad, Flow has been tested, and on select models and hardware continue to improve performance up to 128 cores. However, giving a good overview of the limitations and various factors impacting scaling is a small paper in itself. Nobody has stepped up to provide models with a large number of cells yet. I have not looked at the ONEM models, but you can probably get flow running on them with a little adjustment of the input deck. Just consult the flow manual to see what is supported in flow. Good luck! Alf ________________________________ Fra: Opm på vegne av M. S. Sendt: tirsdag 18. desember 2018 22.09 Til: opm at opm-project.org Emne: [Opm] Large Models for OPM Flow Dear All, I was wondering what's the largest number of processing cores that Flow has been tested with for the MPI parallel version, also if there's a repository having large models (e.g. 1 million cells or larger) ready to test Flow with for benchmarking on 16, 32, 64,128, etc cores. I am currently using the Norne and SPE9 models for testing but they're too small to be run on larger number of cores. I was also previously told here that the larger SPE10 model 2 may not be the best model to use for benchmarking as it's currently not included in any of the integration testing and not the most stable to run. For example ECLIPSE provides a model ONEM1 with different variations for benchmarking on different number of cores (e.g. ONEM1_8, ONEM1_16, etc...), I tried running those models using Flow but it seems that the model input syntax for ECLIPSE isn't 100% compatible with Flow. Any help on this would be much appreciated. Many thanks. Sincerely, Mohamad MIT Geonumerics group ------------------------------------------------------------------- The information contained in this message may be CONFIDENTIAL and is intended for the addressee only. Any unauthorized use, dissemination of the information or copying of this message is prohibited. If you are not the addressee, please notify the sender immediately by return e-mail and delete this message. Thank you From felipe at portella.com.br Wed Dec 5 09:34:49 2018 From: felipe at portella.com.br (Felipe Portella) Date: Wed, 5 Dec 2018 10:34:49 +0100 Subject: [Opm] SPE10 Model 2 is already supported by flow? Message-ID: Hi! I saw a old discussion (August-2017) that flow does not support SPE10 Model 2 at the time, but some improvements for two-phase flows were being incorporated: https://opm-project.org/pipermail/opm/2017-August/001762.html I tried to run the case with OPM Flow 2018.10 with 16 cores on a Xeon but after 14 hours it was still at time 140 of 2000 days. I was wondering if flow already fully supports this model and if this is just a case of tunning of the model to the simulator or if still lacks good preconditioners to this case and that's why is being so slow. If fully supported, what is the average runtime of this model? Thank you! Felipe Portella From rohith.nair at tno.nl Wed Dec 19 16:07:53 2018 From: rohith.nair at tno.nl (Nair, R. (Rohith)) Date: Wed, 19 Dec 2018 16:07:53 +0000 Subject: [Opm] OPM summit 2019 in Utrecht Message-ID: <9ff244235ebe47ae96e6955f018ee0e5@tno.nl> Dear OPM community, We would like to invite you to participate in the 2019 OPM Summit on January 24th and 25th in The Netherlands! This year the event will be hosted by TNO and we hope to welcome all of you in Utrecht. The details of the event are as follows: * Location: Princetonlaan 6, Utrecht, The Netherlands * Dates/Time: 24th and 25th of January 2019. Start: 10:00 on Thursday and end at 15:00 on Friday * Fees: There are no fees for the meeting, Conference facilities and food will be provided by TNO Please let me (rohith.nair at tno.nl) or Frank (frank.wilschut at tno.nl) know as early as possible if you are planning to join the summit, so that we can make the necessary arrangements. We plan to make Thursday (24th) more user centered and Friday (25th) more developer focused, however the agenda is flexible and you can participate on either or both days. The agenda and the presentation slots for the meeting are still open so you are welcome to propose items you would like to share with the rest of the OPM community. If you would like to make a presentation please make the same request to the organizers with a title, authors and a paragraph describing the talk, preferably before 15 January. The agenda will be published on the website and to the participants as soon as it has been set. There will also be a conference dinner on the evening of the 24th (Thursday); in your reply please indicate if you would like to join the dinner. Travel/Accommodation: Participants must make their own arrangements. Our office is on the university campus, but we recommend hotels in the city center or near the central train station (e.g. the NH and Apollo hotels). From the city center, TNO it is easily accessible by bus or taxi. Looking forward to seeing you here! The Organizers (Rohith Nair and Frank Wilschut) This message may contain information that is not intended for you. If you are not the addressee or if this message was sent to you by mistake, you are requested to inform the sender and delete the message. TNO accepts no liability for the content of this e-mail, for the manner in which you use it and for damage of any kind resulting from the risks inherent to the electronic transmission of messages.