- Release-team - mailman3.cels.anl.gov

Re: [Release-team] Next Steps in the build
by Murphy-Olson, Daniel E. 21 Jan '14

21 Jan '14

Agreed, we should try to improve this process, including the spreadsheet. I'm not sure how or why someone showed up as anonymous - we should adjust the permissions on the spreadsheet to only include listed users (no links). For some reason, this morning my dan(a)olsonsolutions.net account doesn't have write access to the spreadsheet - can you fix that as well? The workspace service got special consideration because it needed to be in production in a restricted fashion to migrate data from the old service to the new service. Because of this shock needed to be updated in production as a dependency. Shock's documentation, and release notes hadn't changed since the last release. I thought you were in the planning meeting for the workspace migration, but that may have been the first day of the build when you weren't there? ---- Daniel Murphy-Olson Systems Administrator Mathematics & Computer Science Division Argonne National Laboratory 630-252-0055 ________________________________________ From: Land, Miriam L. [landml(a)ornl.gov] Sent: Tuesday, January 21, 2014 11:22 AM To: Murphy-Olson, Daniel E.; Brettin, Thomas S. Cc: Gordon, India S.; release-team(a)kbase.us Subject: Re: [Release-team] Next Steps in the build Dan, You told me that you have a way to test that the links to the API docs were working in production. That will take care of column O. I will do column P. I believe that Elizabeth is responsible for columns Q (adequate description) and R (public release). I would still like to know why workspace testing is still listed as "in process" and yet it was okayed for deployment by 'anonymous' and Shock was okayed for production without a review of documentation and release notes? It's okay to skip steps but shouldn't we at least know who did it and/or why it was done? Miriam From: <Murphy-Olson>, Dan Olson <dolson(a)mcs.anl.gov<mailto:[email protected]>> Date: Tuesday, January 21, 2014 11:28 AM To: Miriam Land <landml(a)ornl.gov<mailto:[email protected]>>, "Brettin, Thomas S." <brettin(a)cels.anl.gov<mailto:[email protected]>> Cc: "Gordon, India S." <Igordon(a)anl.gov<mailto:[email protected]>>, "release-team(a)kbase.us<mailto:[email protected]>" <release-team(a)kbase.us<mailto:[email protected]>> Subject: RE: [Release-team] please deploy id_map Miriam, No, as far as I know the spreadsheet is correct. I'm not sure who does the OPQR steps, but workspace deluxe and shock service are in there and waiting. I think we need to circle back on a lot of these that are marked "not yet". Who is handling the communication on them? ---- Daniel Murphy-Olson Systems Administrator Mathematics & Computer Science Division Argonne National Laboratory 630-252-0055 ________________________________________ From: Land, Miriam L. [landml(a)ornl.gov<mailto:[email protected]>] Sent: Tuesday, January 21, 2014 9:28 AM To: Brettin, Thomas S.; Murphy-Olson, Daniel E. Cc: Gordon, India S.; release-team(a)kbase.us<mailto:[email protected]> Subject: Re: [Release-team] please deploy id_map Have we completely abandoned the spreadsheet from the last build? Is there any tracking of what is being put into production? How will I know which modules need updated —help so I can send this to Lynn? Miriam From: <Brettin>, "Thomas S." <brettin(a)cels.anl.gov<mailto:[email protected]><mailto:[email protected]>> Date: Monday, January 20, 2014 1:07 PM To: Dan Olson <dolson(a)mcs.anl.gov<mailto:[email protected]><mailto:[email protected]>> Cc: "Gordon, India S." <Igordon(a)anl.gov<mailto:[email protected]><mailto:[email protected]>>, "release-team(a)kbase.us<mailto:[email protected]><mailto:[email protected]>" <release-team(a)kbase.us<mailto:[email protected]><mailto:[email protected]>> Subject: [Release-team] please deploy id_map To production. Please contact me with any errors. Thanks, Tom _______________________________________________ Release-team mailing list Release-team(a)lists.kbase.us<mailto:[email protected]><mailto:[email protected]> https://lists.kbase.us/mailman/listinfo/release-team

1 0

Re: [Release-team] Next Steps in the build
by Land, Miriam L. 21 Jan '14

21 Jan '14

Dan, You told me that you have a way to test that the links to the API docs were working in production. That will take care of column O. I will do column P. I believe that Elizabeth is responsible for columns Q (adequate description) and R (public release). I would still like to know why workspace testing is still listed as "in process" and yet it was okayed for deployment by 'anonymous' and Shock was okayed for production without a review of documentation and release notes? It's okay to skip steps but shouldn't we at least know who did it and/or why it was done? Miriam From: <Murphy-Olson>, Dan Olson <dolson(a)mcs.anl.gov<mailto:[email protected]>> Date: Tuesday, January 21, 2014 11:28 AM To: Miriam Land <landml(a)ornl.gov<mailto:[email protected]>>, "Brettin, Thomas S." <brettin(a)cels.anl.gov<mailto:[email protected]>> Cc: "Gordon, India S." <Igordon(a)anl.gov<mailto:[email protected]>>, "release-team(a)kbase.us<mailto:[email protected]>" <release-team(a)kbase.us<mailto:[email protected]>> Subject: RE: [Release-team] please deploy id_map Miriam, No, as far as I know the spreadsheet is correct. I'm not sure who does the OPQR steps, but workspace deluxe and shock service are in there and waiting. I think we need to circle back on a lot of these that are marked "not yet". Who is handling the communication on them? ---- Daniel Murphy-Olson Systems Administrator Mathematics & Computer Science Division Argonne National Laboratory 630-252-0055 ________________________________________ From: Land, Miriam L. [landml(a)ornl.gov<mailto:[email protected]>] Sent: Tuesday, January 21, 2014 9:28 AM To: Brettin, Thomas S.; Murphy-Olson, Daniel E. Cc: Gordon, India S.; release-team(a)kbase.us<mailto:[email protected]> Subject: Re: [Release-team] please deploy id_map Have we completely abandoned the spreadsheet from the last build? Is there any tracking of what is being put into production? How will I know which modules need updated —help so I can send this to Lynn? Miriam From: <Brettin>, "Thomas S." <brettin(a)cels.anl.gov<mailto:[email protected]><mailto:[email protected]>> Date: Monday, January 20, 2014 1:07 PM To: Dan Olson <dolson(a)mcs.anl.gov<mailto:[email protected]><mailto:[email protected]>> Cc: "Gordon, India S." <Igordon(a)anl.gov<mailto:[email protected]><mailto:[email protected]>>, "release-team(a)kbase.us<mailto:[email protected]><mailto:[email protected]>" <release-team(a)kbase.us<mailto:[email protected]><mailto:[email protected]>> Subject: [Release-team] please deploy id_map To production. Please contact me with any errors. Thanks, Tom _______________________________________________ Release-team mailing list Release-team(a)lists.kbase.us<mailto:[email protected]><mailto:[email protected]> https://lists.kbase.us/mailman/listinfo/release-team

1 0

Re: [Release-team] please deploy id_map
by Robert Olson 21 Jan '14

21 Jan '14

This is excellent news - I'm glad folks (other than me :-) are worrying about the non-root deployments. This all helps among other things with the Mac DMG release. --bob On Jan 21, 2014, at 10:43 AM, Shane Canon wrote: > > Just an update on what I've been working on… > > I've been concentrating on getting all of the software to build as non-root and deploying it on the test cluster. At this time > I have all of the services deployed and starting on the test cluster. I have confirmed that that at least the service starts without spewing errors. > > I've only been able to test a handful of the modules. So it would be best to go through and methodically run the tests to insure everything is working. > > I've committed some changes back to certain modules to make them build cleanly, but I still have some changes that still need to be committed. > > I have also been installing the runtime and deployment at NERSC. So this has given me another datapoint and helped me generalize some fixes. > > What's left: > > We need to methodically test the test deployment. Once that is clean we can rsync the runtime and deployment over to the production cluster. Theoretically > we should be able to just restart services (modula a fixed up deployment.cfg and may some other minor config changes). > > > > > --Shane > > > > > On Jan 21, 2014, at 8:28 AM, "Murphy-Olson, Daniel E." <dolson(a)mcs.anl.gov> wrote: > >> Miriam, >> >> No, as far as I know the spreadsheet is correct. I'm not sure who does the OPQR steps, but workspace deluxe and shock service are in there and waiting. >> >> I think we need to circle back on a lot of these that are marked "not yet". Who is handling the communication on them? >> >> >> ---- >> Daniel Murphy-Olson >> Systems Administrator >> Mathematics & Computer Science Division >> Argonne National Laboratory >> 630-252-0055 >> >> ________________________________________ >> From: Land, Miriam L. [landml(a)ornl.gov] >> Sent: Tuesday, January 21, 2014 9:28 AM >> To: Brettin, Thomas S.; Murphy-Olson, Daniel E. >> Cc: Gordon, India S.; release-team(a)kbase.us >> Subject: Re: [Release-team] please deploy id_map >> >> Have we completely abandoned the spreadsheet from the last build? Is there any tracking of what is being put into production? How will I know which modules need updated —help so I can send this to Lynn? >> >> Miriam >> >> From: <Brettin>, "Thomas S." <brettin(a)cels.anl.gov<mailto:[email protected]>> >> Date: Monday, January 20, 2014 1:07 PM >> To: Dan Olson <dolson(a)mcs.anl.gov<mailto:[email protected]>> >> Cc: "Gordon, India S." <Igordon(a)anl.gov<mailto:[email protected]>>, "release-team(a)kbase.us<mailto:[email protected]>" <release-team(a)kbase.us<mailto:[email protected]>> >> Subject: [Release-team] please deploy id_map >> >> To production. >> >> Please contact me with any errors. >> >> Thanks, >> Tom >> _______________________________________________ >> Release-team mailing list >> Release-team(a)lists.kbase.us<mailto:[email protected]> >> https://lists.kbase.us/mailman/listinfo/release-team >> >> _______________________________________________ >> Release-team mailing list >> Release-team(a)lists.kbase.us >> https://lists.kbase.us/mailman/listinfo/release-team > > _______________________________________________ > Release-team mailing list > Release-team(a)lists.kbase.us > https://lists.kbase.us/mailman/listinfo/release-team

1 0

Re: [Release-team] please deploy id_map
by Shane Canon 21 Jan '14

21 Jan '14

Just an update on what I've been working on… I've been concentrating on getting all of the software to build as non-root and deploying it on the test cluster. At this time I have all of the services deployed and starting on the test cluster. I have confirmed that that at least the service starts without spewing errors. I've only been able to test a handful of the modules. So it would be best to go through and methodically run the tests to insure everything is working. I've committed some changes back to certain modules to make them build cleanly, but I still have some changes that still need to be committed. I have also been installing the runtime and deployment at NERSC. So this has given me another datapoint and helped me generalize some fixes. What's left: We need to methodically test the test deployment. Once that is clean we can rsync the runtime and deployment over to the production cluster. Theoretically we should be able to just restart services (modula a fixed up deployment.cfg and may some other minor config changes). --Shane On Jan 21, 2014, at 8:28 AM, "Murphy-Olson, Daniel E." <dolson(a)mcs.anl.gov> wrote: > Miriam, > > No, as far as I know the spreadsheet is correct. I'm not sure who does the OPQR steps, but workspace deluxe and shock service are in there and waiting. > > I think we need to circle back on a lot of these that are marked "not yet". Who is handling the communication on them? > > > ---- > Daniel Murphy-Olson > Systems Administrator > Mathematics & Computer Science Division > Argonne National Laboratory > 630-252-0055 > > ________________________________________ > From: Land, Miriam L. [landml(a)ornl.gov] > Sent: Tuesday, January 21, 2014 9:28 AM > To: Brettin, Thomas S.; Murphy-Olson, Daniel E. > Cc: Gordon, India S.; release-team(a)kbase.us > Subject: Re: [Release-team] please deploy id_map > > Have we completely abandoned the spreadsheet from the last build? Is there any tracking of what is being put into production? How will I know which modules need updated —help so I can send this to Lynn? > > Miriam > > From: <Brettin>, "Thomas S." <brettin(a)cels.anl.gov<mailto:[email protected]>> > Date: Monday, January 20, 2014 1:07 PM > To: Dan Olson <dolson(a)mcs.anl.gov<mailto:[email protected]>> > Cc: "Gordon, India S." <Igordon(a)anl.gov<mailto:[email protected]>>, "release-team(a)kbase.us<mailto:[email protected]>" <release-team(a)kbase.us<mailto:[email protected]>> > Subject: [Release-team] please deploy id_map > > To production. > > Please contact me with any errors. > > Thanks, > Tom > _______________________________________________ > Release-team mailing list > Release-team(a)lists.kbase.us<mailto:[email protected]> > https://lists.kbase.us/mailman/listinfo/release-team > > _______________________________________________ > Release-team mailing list > Release-team(a)lists.kbase.us > https://lists.kbase.us/mailman/listinfo/release-team

1 0

Re: [Release-team] please deploy id_map
by Murphy-Olson, Daniel E. 21 Jan '14

21 Jan '14

Miriam, No, as far as I know the spreadsheet is correct. I'm not sure who does the OPQR steps, but workspace deluxe and shock service are in there and waiting. I think we need to circle back on a lot of these that are marked "not yet". Who is handling the communication on them? ---- Daniel Murphy-Olson Systems Administrator Mathematics & Computer Science Division Argonne National Laboratory 630-252-0055 ________________________________________ From: Land, Miriam L. [landml(a)ornl.gov] Sent: Tuesday, January 21, 2014 9:28 AM To: Brettin, Thomas S.; Murphy-Olson, Daniel E. Cc: Gordon, India S.; release-team(a)kbase.us Subject: Re: [Release-team] please deploy id_map Have we completely abandoned the spreadsheet from the last build? Is there any tracking of what is being put into production? How will I know which modules need updated —help so I can send this to Lynn? Miriam From: <Brettin>, "Thomas S." <brettin(a)cels.anl.gov<mailto:[email protected]>> Date: Monday, January 20, 2014 1:07 PM To: Dan Olson <dolson(a)mcs.anl.gov<mailto:[email protected]>> Cc: "Gordon, India S." <Igordon(a)anl.gov<mailto:[email protected]>>, "release-team(a)kbase.us<mailto:[email protected]>" <release-team(a)kbase.us<mailto:[email protected]>> Subject: [Release-team] please deploy id_map To production. Please contact me with any errors. Thanks, Tom _______________________________________________ Release-team mailing list Release-team(a)lists.kbase.us<mailto:[email protected]> https://lists.kbase.us/mailman/listinfo/release-team

1 0

Re: [Release-team] please deploy id_map
by Land, Miriam L. 21 Jan '14

21 Jan '14

Have we completely abandoned the spreadsheet from the last build? Is there any tracking of what is being put into production? How will I know which modules need updated —help so I can send this to Lynn? Miriam From: <Brettin>, "Thomas S." <brettin(a)cels.anl.gov<mailto:[email protected]>> Date: Monday, January 20, 2014 1:07 PM To: Dan Olson <dolson(a)mcs.anl.gov<mailto:[email protected]>> Cc: "Gordon, India S." <Igordon(a)anl.gov<mailto:[email protected]>>, "release-team(a)kbase.us<mailto:[email protected]>" <release-team(a)kbase.us<mailto:[email protected]>> Subject: [Release-team] please deploy id_map To production. Please contact me with any errors. Thanks, Tom _______________________________________________ Release-team mailing list Release-team(a)lists.kbase.us<mailto:[email protected]> https://lists.kbase.us/mailman/listinfo/release-team

1 0

please deploy id_map
by Brettin, Thomas S. 20 Jan '14

20 Jan '14

To production. Please contact me with any errors. Thanks, Tom

1 0

Re: [Release-team] Python servers non-functional in January build
by Murphy-Olson, Daniel E. 20 Jan '14

20 Jan '14

t1 was used for most of the testing, and some of the testing requires the service to be running on the local host. Feel free to shut down those services and run them on the dedicated hosts if necessary. ---- Daniel Murphy-Olson Systems Administrator Mathematics & Computer Science Division Argonne National Laboratory 630-252-0055<callto:630-252-0055> ________________________________ From: release-team-bounces(a)lists.kbase.us [release-team-bounces(a)lists.kbase.us] on behalf of Mundy, Michael B. (Mike) [Mundy.Michael(a)mayo.edu] Sent: Monday, January 20, 2014 9:26 AM To: Shane Canon Cc: Chia, Nicholas, Ph.D.; release-team(a)kbase.us; Henry, Christopher S.; Matthew Benedict; Stephen Chan; N.L. Harris Subject: Re: [Release-team] Python servers non-functional in January build Shane, Thanks for adding cogent. I forgot that cogent has a dependency on mpi4py (it throws annoying error messages if it is not found). I was running on t1. I'm confused by the new test environment for this build. Are all of the servers supposed to be running on t1? Or on the instances specified in /root/services? For example, the services file says that workspace_deluxe should be running on t49 at port 7058. But when I checked on t49 this morning there was nothing listening on port 7058. I'll be at Argonne Tuesday and Wednesday. Who would be the best person to work with there? Mike Mundy | Senior Analyst Programmer | Bioinformatics Systems Unit | 507-266-7232 | mundy.michael(a)mayo.edu<mailto:[email protected]> Mayo Clinic | 200 First Street SW | Rochester, MN 55905 | mayoclinic.org<http://www.mayoclinic.org> From: Shane Canon <scanon(a)lbl.gov<mailto:[email protected]>> Date: Friday, January 17, 2014 2:39 PM To: Michael Mundy <mundy.michael(a)mayo.edu<mailto:[email protected]>> Cc: "release-team(a)kbase.us<mailto:[email protected]>" <release-team(a)kbase.us<mailto:[email protected]>>, Christopher Henry <chrisshenry(a)gmail.com<mailto:[email protected]>>, Stephen Chan <sychan(a)lbl.gov<mailto:[email protected]>>, Matthew Benedict <matthew.n.benedict(a)gmail.com<mailto:[email protected]>>, "Chia, Nicholas, Ph.D." <Chia.Nicholas(a)mayo.edu<mailto:[email protected]>>, "N.L. Harris" <nlharris(a)lbl.gov<mailto:[email protected]>> Subject: Re: [Release-team] Python servers non-functional in January build 1. I'm not sure about this. Is this on t1 or production? It could be the client tools aren't matched. ws is the new workspace which isn't "public" yet. 2. I'm seeing this too and I think it is related to going through the nginx proxy. There have been some discussions about this. 3. I added cogent on the 2014 runtime on t1. Let me know if you need more. --Shane On Jan 17, 2014, at 10:45 AM, "Mundy, Michael B. (Mike)" <Mundy.Michael(a)mayo.edu<mailto:[email protected]>> wrote: Shane, Here's what I found in my testing: 1. I used https://kbase.us/services/ws/ as the URL for the new workspace service but I'm getting a "HTTPError: HTTP Error 502: Bad Gateway" error when accessing it. 2. I'm getting a "client-tests/1_V2.fasta: error: Error creating node: request error: 411 Length Required" error when trying to upload a file to shock. The shock team had already looked at this and said it was a problem with the old shock server. The shock-client.cfg file for the ubuntu user has the server URL set to url=http://kbase.us/services/shock-api. Is that URL going to the server in the new deployment. 3. The cogent Python package isn't installed in the kbase Python installation (cogent is installed in the system Python installation). Mike Mundy | Senior Analyst Programmer | Bioinformatics Systems Unit | 507-266-7232 | mundy.michael(a)mayo.edu<mailto:[email protected]> Mayo Clinic | 200 First Street SW | Rochester, MN 55905 | mayoclinic.org<http://www.mayoclinic.org/> From: Shane Canon <scanon(a)lbl.gov<mailto:[email protected]>> Date: Friday, January 17, 2014 11:01 AM To: Michael Mundy <mundy.michael(a)mayo.edu<mailto:[email protected]>> Cc: "release-team(a)kbase.us<mailto:[email protected]>" <release-team(a)kbase.us<mailto:[email protected]>>, Christopher Henry <chrisshenry(a)gmail.com<mailto:[email protected]>>, Stephen Chan <sychan(a)lbl.gov<mailto:[email protected]>>, Matthew Benedict <matthew.n.benedict(a)gmail.com<mailto:[email protected]>>, "Chia, Nicholas, Ph.D." <Chia.Nicholas(a)mayo.edu<mailto:[email protected]>>, "N.L. Harris" <nlharris(a)lbl.gov<mailto:[email protected]>> Subject: Re: [Release-team] Python servers non-functional in January build Mike, When I originally built the runtime and deployment, I inadvertently used 2013 instead of 2014 in the path. I've created a new runtime and deployment with the correct date (/kbase/deployments/20140109-prod/ and /kbase/runtimes/20140109-prod/). That version should have Steve's changes. Can you test that out? I still need to restart a few services that were originally started with the 2013 tree. I will do that now. --Shane On Jan 17, 2014, at 8:43 AM, "Mundy, Michael B. (Mike)" <Mundy.Michael(a)mayo.edu<mailto:[email protected]>> wrote: Authenticated Python servers are currently non-functional in the January build. Steve Chan has fixed the problem with changes to the auth module with the following commits: commit 12f346a4ed8e2d30a580d85e7e7987c78568b9bb Merge: 1d98332 0eca1d4 Author: Steve Chan <sychan(a)lbl.gov<mailto:[email protected]>> Date: Mon Jan 13 15:46:37 2014 -0800 Merge branch 'master' into dev_requests-1.x Conflicts: Bio-KBase-Auth/lib/Bio/KBase/Auth.pm python-libs/biokbase/nexus/__init__.py commit 0eca1d43361b8955b0d9ce391778cd5a86561150 Author: Steve Chan <sychan(a)lbl.gov<mailto:[email protected]>> Date: Mon Jan 13 15:41:46 2014 -0800 Part of cleanup for the Nexus/requests library API drift Can we get these two commits in the build and installed on the t1 system. The Probabilistic Annotation and Compression Based Distance services cannot complete the build until this is done. Mike Mundy | Senior Analyst Programmer | Bioinformatics Systems Unit | 507-266-7232 | mundy.michael(a)mayo.edu<mailto:[email protected]> Mayo Clinic | 200 First Street SW | Rochester, MN 55905 | mayoclinic.org<http://www.mayoclinic.org/> _______________________________________________ Release-team mailing list Release-team(a)lists.kbase.us<mailto:[email protected]> https://lists.kbase.us/mailman/listinfo/release-team

1 0

Re: [Release-team] Python servers non-functional in January build
by Mundy, Michael B. (Mike) 20 Jan '14

20 Jan '14

Shane, Thanks for adding cogent. I forgot that cogent has a dependency on mpi4py (it throws annoying error messages if it is not found). I was running on t1. I'm confused by the new test environment for this build. Are all of the servers supposed to be running on t1? Or on the instances specified in /root/services? For example, the services file says that workspace_deluxe should be running on t49 at port 7058. But when I checked on t49 this morning there was nothing listening on port 7058. I'll be at Argonne Tuesday and Wednesday. Who would be the best person to work with there? Mike Mundy | Senior Analyst Programmer | Bioinformatics Systems Unit | 507-266-7232 | mundy.michael(a)mayo.edu<mailto:[email protected]> Mayo Clinic | 200 First Street SW | Rochester, MN 55905 | mayoclinic.org<http://www.mayoclinic.org> From: Shane Canon <scanon(a)lbl.gov<mailto:[email protected]>> Date: Friday, January 17, 2014 2:39 PM To: Michael Mundy <mundy.michael(a)mayo.edu<mailto:[email protected]>> Cc: "release-team(a)kbase.us<mailto:[email protected]>" <release-team(a)kbase.us<mailto:[email protected]>>, Christopher Henry <chrisshenry(a)gmail.com<mailto:[email protected]>>, Stephen Chan <sychan(a)lbl.gov<mailto:[email protected]>>, Matthew Benedict <matthew.n.benedict(a)gmail.com<mailto:[email protected]>>, "Chia, Nicholas, Ph.D." <Chia.Nicholas(a)mayo.edu<mailto:[email protected]>>, "N.L. Harris" <nlharris(a)lbl.gov<mailto:[email protected]>> Subject: Re: [Release-team] Python servers non-functional in January build 1. I'm not sure about this. Is this on t1 or production? It could be the client tools aren't matched. ws is the new workspace which isn't "public" yet. 2. I'm seeing this too and I think it is related to going through the nginx proxy. There have been some discussions about this. 3. I added cogent on the 2014 runtime on t1. Let me know if you need more. --Shane On Jan 17, 2014, at 10:45 AM, "Mundy, Michael B. (Mike)" <Mundy.Michael(a)mayo.edu<mailto:[email protected]>> wrote: Shane, Here's what I found in my testing: 1. I used https://kbase.us/services/ws/ as the URL for the new workspace service but I'm getting a "HTTPError: HTTP Error 502: Bad Gateway" error when accessing it. 2. I'm getting a "client-tests/1_V2.fasta: error: Error creating node: request error: 411 Length Required" error when trying to upload a file to shock. The shock team had already looked at this and said it was a problem with the old shock server. The shock-client.cfg file for the ubuntu user has the server URL set to url=http://kbase.us/services/shock-api. Is that URL going to the server in the new deployment. 3. The cogent Python package isn't installed in the kbase Python installation (cogent is installed in the system Python installation). Mike Mundy | Senior Analyst Programmer | Bioinformatics Systems Unit | 507-266-7232 | mundy.michael(a)mayo.edu<mailto:[email protected]> Mayo Clinic | 200 First Street SW | Rochester, MN 55905 | mayoclinic.org<http://www.mayoclinic.org/> From: Shane Canon <scanon(a)lbl.gov<mailto:[email protected]>> Date: Friday, January 17, 2014 11:01 AM To: Michael Mundy <mundy.michael(a)mayo.edu<mailto:[email protected]>> Cc: "release-team(a)kbase.us<mailto:[email protected]>" <release-team(a)kbase.us<mailto:[email protected]>>, Christopher Henry <chrisshenry(a)gmail.com<mailto:[email protected]>>, Stephen Chan <sychan(a)lbl.gov<mailto:[email protected]>>, Matthew Benedict <matthew.n.benedict(a)gmail.com<mailto:[email protected]>>, "Chia, Nicholas, Ph.D." <Chia.Nicholas(a)mayo.edu<mailto:[email protected]>>, "N.L. Harris" <nlharris(a)lbl.gov<mailto:[email protected]>> Subject: Re: [Release-team] Python servers non-functional in January build Mike, When I originally built the runtime and deployment, I inadvertently used 2013 instead of 2014 in the path. I've created a new runtime and deployment with the correct date (/kbase/deployments/20140109-prod/ and /kbase/runtimes/20140109-prod/). That version should have Steve's changes. Can you test that out? I still need to restart a few services that were originally started with the 2013 tree. I will do that now. --Shane On Jan 17, 2014, at 8:43 AM, "Mundy, Michael B. (Mike)" <Mundy.Michael(a)mayo.edu<mailto:[email protected]>> wrote: Authenticated Python servers are currently non-functional in the January build. Steve Chan has fixed the problem with changes to the auth module with the following commits: commit 12f346a4ed8e2d30a580d85e7e7987c78568b9bb Merge: 1d98332 0eca1d4 Author: Steve Chan <sychan(a)lbl.gov<mailto:[email protected]>> Date: Mon Jan 13 15:46:37 2014 -0800 Merge branch 'master' into dev_requests-1.x Conflicts: Bio-KBase-Auth/lib/Bio/KBase/Auth.pm python-libs/biokbase/nexus/__init__.py commit 0eca1d43361b8955b0d9ce391778cd5a86561150 Author: Steve Chan <sychan(a)lbl.gov<mailto:[email protected]>> Date: Mon Jan 13 15:41:46 2014 -0800 Part of cleanup for the Nexus/requests library API drift Can we get these two commits in the build and installed on the t1 system. The Probabilistic Annotation and Compression Based Distance services cannot complete the build until this is done. Mike Mundy | Senior Analyst Programmer | Bioinformatics Systems Unit | 507-266-7232 | mundy.michael(a)mayo.edu<mailto:[email protected]> Mayo Clinic | 200 First Street SW | Rochester, MN 55905 | mayoclinic.org<http://www.mayoclinic.org/> _______________________________________________ Release-team mailing list Release-team(a)lists.kbase.us<mailto:[email protected]> https://lists.kbase.us/mailman/listinfo/release-team

1 0

Re: [Release-team] Python servers non-functional in January build
by Shane Canon 17 Jan '14

17 Jan '14

1. I'm not sure about this. Is this on t1 or production? It could be the client tools aren't matched. ws is the new workspace which isn't "public" yet. 2. I'm seeing this too and I think it is related to going through the nginx proxy. There have been some discussions about this. 3. I added cogent on the 2014 runtime on t1. Let me know if you need more. --Shane On Jan 17, 2014, at 10:45 AM, "Mundy, Michael B. (Mike)" <Mundy.Michael(a)mayo.edu> wrote: > Shane, > > Here's what I found in my testing: > I used https://kbase.us/services/ws/ as the URL for the new workspace service but I'm getting a "HTTPError: HTTP Error 502: Bad Gateway" error when accessing it. > I'm getting a "client-tests/1_V2.fasta: error: Error creating node: request error: 411 Length Required" error when trying to upload a file to shock. The shock team had already looked at this and said it was a problem with the old shock server. The shock-client.cfg file for the ubuntu user has the server URL set to url=http://kbase.us/services/shock-api. Is that URL going to the server in the new deployment. > The cogent Python package isn't installed in the kbase Python installation (cogent is installed in the system Python installation). > Mike Mundy | Senior Analyst Programmer | Bioinformatics Systems Unit | 507-266-7232 | mundy.michael(a)mayo.edu > Mayo Clinic | 200 First Street SW | Rochester, MN 55905 | mayoclinic.org > > From: Shane Canon <scanon(a)lbl.gov> > Date: Friday, January 17, 2014 11:01 AM > To: Michael Mundy <mundy.michael(a)mayo.edu> > Cc: "release-team(a)kbase.us" <release-team(a)kbase.us>, Christopher Henry <chrisshenry(a)gmail.com>, Stephen Chan <sychan(a)lbl.gov>, Matthew Benedict <matthew.n.benedict(a)gmail.com>, "Chia, Nicholas, Ph.D." <Chia.Nicholas(a)mayo.edu>, "N.L. Harris" <nlharris(a)lbl.gov> > Subject: Re: [Release-team] Python servers non-functional in January build > > > Mike, > > When I originally built the runtime and deployment, I inadvertently used 2013 instead of 2014 in the path. I've created a new runtime and deployment with the correct date (/kbase/deployments/20140109-prod/ and > /kbase/runtimes/20140109-prod/). That version should have Steve's changes. Can you test that out? I still need to restart a few services that were originally started with the 2013 tree. I will do that now. > > > --Shane > > On Jan 17, 2014, at 8:43 AM, "Mundy, Michael B. (Mike)" <Mundy.Michael(a)mayo.edu> wrote: > >> Authenticated Python servers are currently non-functional in the January build. Steve Chan has fixed the problem with changes to the auth module with the following commits: >> >> commit 12f346a4ed8e2d30a580d85e7e7987c78568b9bb >> Merge: 1d98332 0eca1d4 >> Author: Steve Chan <sychan(a)lbl.gov> >> Date: Mon Jan 13 15:46:37 2014 -0800 >> >> Merge branch 'master' into dev_requests-1.x >> >> Conflicts: >> Bio-KBase-Auth/lib/Bio/KBase/Auth.pm >> python-libs/biokbase/nexus/__init__.py >> >> commit 0eca1d43361b8955b0d9ce391778cd5a86561150 >> Author: Steve Chan <sychan(a)lbl.gov> >> Date: Mon Jan 13 15:41:46 2014 -0800 >> >> Part of cleanup for the Nexus/requests library API drift >> >> Can we get these two commits in the build and installed on the t1 system. The Probabilistic Annotation and Compression Based Distance services cannot complete the build until this is done. >> >> Mike Mundy | Senior Analyst Programmer | Bioinformatics Systems Unit | 507-266-7232 | mundy.michael(a)mayo.edu >> Mayo Clinic | 200 First Street SW | Rochester, MN 55905 | mayoclinic.org >> _______________________________________________ >> Release-team mailing list >> Release-team(a)lists.kbase.us >> https://lists.kbase.us/mailman/listinfo/release-team >

1 0