Open curtisecombsjr opened 9 months ago
Strange, but I enabled "log_level: debug" to see if i could find an error, but that seems to have broken it even more. Now the pav.log does not show any output and the series list is completely empty...
~> pav list series
No matching items found.
~>
Also a we are running 2.3 if that helps.
:~> pav --version
Pavilion 2.3
:~>
Thanks
Sorry for the delay - we've been out the last few weeks for holiday break.
Double check that working directory path Pavilion is using is what you think it is.
pav show config
will show you the Pavilion config settings as Pavilion sees them. Double check the path as given there.
Also, pav show config_dir
will list all the config directories.
Paul,
Thanks so much. Everything in the output here seems to be correct. Let me attach it to this comment, so that you can see for yourself, but as far as I can tell, the directories line up and match. New mystery: pav list series
now shows no series. Very strange.
pav.txt
Are you using the latest master?
Could you attach you pavilion.yaml?
From: curtisecombsjr @.> Sent: Tuesday, January 2, 2024 12:38 PM To: hpc/pavilion2 @.> Cc: Ferrell, Paul Steven @.>; Comment @.> Subject: [EXTERNAL] Re: [hpc/pavilion2] older series are not show (Issue #722)
Paul,
Thanks so much. Everything in the output here seems to be correct. Let me attach it to this comment, so that you can see for yourself, but as far as I can tell, the directories line up and match. New mystery: pav list series now shows series. Very strange. pav.txthttps://urldefense.com/v3/__https://github.com/hpc/pavilion2/files/13813370/pav.txt__;!!Bt8fGhp8LhKGRg!AngOimVYB2gfZga0ZDO9FWrkhANSwrt3J4bmqx-xT6HT9FYS-FnLNU676GR5OWieI2fKZPuvatKu0_IOY7hif6zrHA$
— Reply to this email directly, view it on GitHubhttps://urldefense.com/v3/__https://github.com/hpc/pavilion2/issues/722*issuecomment-1874462589__;Iw!!Bt8fGhp8LhKGRg!AngOimVYB2gfZga0ZDO9FWrkhANSwrt3J4bmqx-xT6HT9FYS-FnLNU676GR5OWieI2fKZPuvatKu0_IOY7jMC1anvA$, or unsubscribehttps://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AMK6D5BIT4ENWAOGKQKQZMTYMRO4JAVCNFSM6AAAAABA7F5MFOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNZUGQ3DENJYHE__;!!Bt8fGhp8LhKGRg!AngOimVYB2gfZga0ZDO9FWrkhANSwrt3J4bmqx-xT6HT9FYS-FnLNU676GR5OWieI2fKZPuvatKu0_IOY7iXWrRyUw$. You are receiving this because you commented.Message ID: @.***>
I tried to switch to the latest clone that I could, but unfortunately, our nodes aren't fully updated and i got errors, so i had to switch back. That being said, I do not think that this version is that old, might be a year-ish. Is there a way that I can check? It would be very difficult for us to upgrade at this moment, either way, sadly. A node OS update won't be for a while and even then we are not updating to the latest SP (Suse) here's my yaml. pav_config.txt
No problem. If you're using a git checkout of Pavilion, just send me the git hash. I can check out the version you're using and work from there.
From: curtisecombsjr @.> Sent: Tuesday, January 2, 2024 1:26 PM To: hpc/pavilion2 @.> Cc: Ferrell, Paul Steven @.>; Comment @.> Subject: [EXTERNAL] Re: [hpc/pavilion2] older series are not show (Issue #722)
I tried to switch to the latest clone that I could, but unfortunately, our nodes aren't fully updated and i got errors, so i had to switch back. That being said, I do not think that this version is that old, might be a year-ish. Is there a way that I can check? It would be very difficult for us to upgrade at this moment, either way, sadly. A node OS update won't be for a while and even then we are not updating to the latest SP (Suse) here's my yaml. pav_config.txthttps://urldefense.com/v3/__https://github.com/hpc/pavilion2/files/13813703/pav_config.txt__;!!Bt8fGhp8LhKGRg!C66X8hDVP-Ls3VPiEC4pWcGX8fd5Dbpzdf8syLuoUj0XjQl2uFvtDwy_RcCYF0X3Ie0_6TdHIVc498RhJP6B-9xEpA$
— Reply to this email directly, view it on GitHubhttps://urldefense.com/v3/__https://github.com/hpc/pavilion2/issues/722*issuecomment-1874510155__;Iw!!Bt8fGhp8LhKGRg!C66X8hDVP-Ls3VPiEC4pWcGX8fd5Dbpzdf8syLuoUj0XjQl2uFvtDwy_RcCYF0X3Ie0_6TdHIVc498RhJP6ww9CHAQ$, or unsubscribehttps://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AMK6D5BVEUUHRQMXJILOR73YMRUPFAVCNFSM6AAAAABA7F5MFOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNZUGUYTAMJVGU__;!!Bt8fGhp8LhKGRg!C66X8hDVP-Ls3VPiEC4pWcGX8fd5Dbpzdf8syLuoUj0XjQl2uFvtDwy_RcCYF0X3Ie0_6TdHIVc498RhJP4UcGDl3w$. You are receiving this because you commented.Message ID: @.***>
Could this be helpful? It's been moved around and I wasn't the original person that installed it. Kinda just inherited it:
hpc-adm@clogin09:/apps/dev/pavilion/src/pavilion2> git log | cat | head -20
commit 6d6e359bc551e3981b14a5516a4028b96e2f3042
Author: Francine Lapid <55203623+francinelapid@users.noreply.github.com>
Date: Tue Apr 6 11:17:44 2021 -0600
fixed command-line overrides (#397)
Co-authored-by: Paul Ferrell <51765748+Paul-Ferrell@users.noreply.github.com>
commit 07e9ebe47b5504132832c7564c97b813508337ca
Author: Francine Lapid <55203623+francinelapid@users.noreply.github.com>
Date: Tue Apr 6 11:05:24 2021 -0600
cancels entire series if there's a scheduler error (#399)
* cancels entire series if there's a scheduler error
* includes test id in output error
* replaced by_sigterm parameter with message
hpc-adm@clogin09:/apps/dev/pavilion/src/pavilion2> git rev-parse --short HEAD
6d6e359b
hpc-adm@clogin09:/apps/dev/pavilion/src/pavilion2>
Yep, that will do it. Give me a bit - I've got other fires to handle, but I'll get to this today.
That version is quite old, and I think I vaguely remember this bug.
From: curtisecombsjr @.> Sent: Tuesday, January 2, 2024 1:52 PM To: hpc/pavilion2 @.> Cc: Ferrell, Paul Steven @.>; Comment @.> Subject: [EXTERNAL] Re: [hpc/pavilion2] older series are not show (Issue #722)
Could this be helpful? It's been moved around and I wasn't the original person that installed it. Kinda just inherited it:
@.:/apps/dev/pavilion/src/pavilion2> git log | cat | head -20 commit 6d6e359https://urldefense.com/v3/__https://github.com/hpc/pavilion2/commit/6d6e359bc551e3981b14a5516a4028b96e2f3042__;!!Bt8fGhp8LhKGRg!GLvTj_AE_Zyl9KmEtfOiOpOmAVD8q47i44IeNIz0nk0HeJfWy5T1qaw8JnLUthMJgX9Ypn-6nNhSvsmM1UeNvNTTJw$ Author: Francine Lapid @*.**@*.***> Date: Tue Apr 6 11:17:44 2021 -0600
fixed command-line overrides (#397)
Co-authored-by: Paul Ferrell @.***>
commit 07e9ebehttps://urldefense.com/v3/__https://github.com/hpc/pavilion2/commit/07e9ebe47b5504132832c7564c97b813508337ca__;!!Bt8fGhp8LhKGRg!GLvTj_AE_Zyl9KmEtfOiOpOmAVD8q47i44IeNIz0nk0HeJfWy5T1qaw8JnLUthMJgX9Ypn-6nNhSvsmM1UeeWvUBiQ$ Author: Francine Lapid @.**@.> Date: Tue Apr 6 11:05:24 2021 -0600
cancels entire series if there's a scheduler error (#399)
cancels entire series if there's a scheduler error
includes test id in output error
replaced by_sigterm parameter with message
@.:/apps/dev/pavilion/src/pavilion2> git rev-parse --short HEAD 6d6e359https://urldefense.com/v3/__https://github.com/hpc/pavilion2/commit/6d6e359bc551e3981b14a5516a4028b96e2f3042__;!!Bt8fGhp8LhKGRg!GLvTj_AE_Zyl9KmEtfOiOpOmAVD8q47i44IeNIz0nk0HeJfWy5T1qaw8JnLUthMJgX9Ypn-6nNhSvsmM1UeNvNTTJw$ @.:/apps/dev/pavilion/src/pavilion2>`
— Reply to this email directly, view it on GitHubhttps://urldefense.com/v3/__https://github.com/hpc/pavilion2/issues/722*issuecomment-1874537285__;Iw!!Bt8fGhp8LhKGRg!GLvTj_AE_Zyl9KmEtfOiOpOmAVD8q47i44IeNIz0nk0HeJfWy5T1qaw8JnLUthMJgX9Ypn-6nNhSvsmM1UfpU4BVRg$, or unsubscribehttps://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AMK6D5CGAKWZV5PGWGGWWVDYMRXRDAVCNFSM6AAAAABA7F5MFOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNZUGUZTOMRYGU__;!!Bt8fGhp8LhKGRg!GLvTj_AE_Zyl9KmEtfOiOpOmAVD8q47i44IeNIz0nk0HeJfWy5T1qaw8JnLUthMJgX9Ypn-6nNhSvsmM1UdGSqLCtw$. You are receiving this because you commented.Message ID: @.***>
Awesome, thank you! And take your time. This is not seriously affecting us at the moment. We run pavilion tests for our burn-ins and it's working fine for those, it would just be a problem if we needed to look at the past results (which does happen, but not that often). Thanks again!
Hello! We have several hundred series in .working_dir/series but pavilion seems to only "know about" the last 2 and outputs the incorrect tests for older series when queried with
pav status s<sid>
I am using "pav status" on an older series, 675. The current series is 788, however the testids for s788 are shown for s675.
(notice here, that the Pavilion testids listed in the 0000675 directory are 3009-3012, NOT 3460-3465)
This only started happening today. Thank you so much!