Re: pkg builders order of jobs
Tatsuki Makino tatsuki_makino at hotmail.com wrote on
Mon May 10 21:08:06 UTC 2021 :
> In poudriere.conf, there is a setting for BUILDNAME_FORMAT.
> It will be set to the default value of %Y-%m-%d_%Hh%Mm%Ss in
> /usr/local/share/poudriere/common.sh
>
> BUILDNAME_FORMAT is used as ${BUILDNAME:=$(date +${BUILDNAME_FORMAT})} and
> becomes the default value for BUILDNAME.
> BUILDNAME can be set with the -B option of poudriere-bulk.
>
> In other words, I want the official poudriere bulk to start with an option
> like -B 210511060619-53863351e3fbec0d89e941ee9f0f288bd2461e87 too :)
Another, shorter option would be to truncate the hash to,
say, 12 characters (and I show all 4 digits of the year):
-B 20210511060619-53863351e3fb
Especially with content like 20210511060619 as context
in the name, 53863351e3fb is unlikely to not be
unique for built-commit identification.
However, the port builds identify two commits: one for
the port (p prefixed currently) and one for the system
(_s prefixed currently). So more like (using a real
build's hashes):
-B 20210511060619-p0ae18b07e470_sf5ff282bc0
Note that the text after the "-" is the same text now
used in the name.
The use of the date/time encoding as a prefix avoids
dealing with picking between the port and system
"rev-list --first-parent --count" figure or having to
list both in some order --and avoids the computational
effort to find those counts.
===
Mark Millard
marklmi at yahoo.com
( dsl-only.net went
away in early 2018-Mar)
___
freebsd-ports@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: pkg builders order of jobs
Ronald Klop wrote on 2021/05/10 18:54:
> At home my Poudriere names the builds like "2021-05-09_20h49m55s". I don't
> remember configuring this by myself, so I think it is the default.
In poudriere.conf, there is a setting for BUILDNAME_FORMAT.
It will be set to the default value of %Y-%m-%d_%Hh%Mm%Ss in
/usr/local/share/poudriere/common.sh
BUILDNAME_FORMAT is used as ${BUILDNAME:=$(date +${BUILDNAME_FORMAT})} and
becomes the default value for BUILDNAME.
BUILDNAME can be set with the -B option of poudriere-bulk.
In other words, I want the official poudriere bulk to start with an option like
-B 210511060619-53863351e3fbec0d89e941ee9f0f288bd2461e87 too :)
___
freebsd-ports@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: pkg builders order of jobs
Van: Stefan Esser Datum: zondag, 9 mei 2021 12:40 Aan: po...@freebsd.org Onderwerp: Re: pkg builders order of jobs Am 09.05.21 um 12:22 schrieb Ronald Klop: > Hi, > > On the pkg builders the list of builds is ordered on git hash. > http://beefy18.nyi.freebsd.org/jail.html?mastername=main-amd64-default > (NB: ipv6 only) > > This ordering is totally random to me. With svn the list was naturally ordered > because of the incremental commit id. > > Is it possible to add the commit count to the job-name like sys/conf/newvers.sh > in base? > git_cnt=$($git_cmd rev-list --first-parent --count HEAD 2>/dev/null) Adding the start-time of the build in ISO format would also be highly appreciated, at least by me ;-) The date should reflect the time of the last commit that has been considered. Such a time stamp would let me see with little effort whether a failure is due to a long running build job that did not catch-up with the latest fixes to a port, or whether there still is an issue with a port (especially for the "exotic" architectures that I do not test on my system and that often need a lot longer to complete a build job). Regards, STefan Hi, Which method is used to create a sortable list for human readabiltiy does not really matter to me. Commit count or unix timestamp or 20210510_1147 or some Poudriere internal build number or something I did not think of. Everything is better than pseudo-random hex strings to me. :-) At home my Poudriere names the builds like "2021-05-09_20h49m55s". I don't remember configuring this by myself, so I think it is the default. But I can imagine people also want an exact reference to the git hash. Maybe the SVN property in the Poudriere page can be used for that. I don't know, somebody with more inside knowledge should decide. Regards, Ronald. ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: pkg builders order of jobs
Ronald Klop ronald-lists at klop.ws wrote on Sun May 9 10:22:38 UTC 2021 : > On the pkg builders the list of builds is ordered on git hash. > > http://beefy18.nyi.freebsd.org/jail.html?mastername=main-amd64-default > > (NB: ipv6 only) > > This ordering is totally random to me. With svn the list was naturally > ordered because of the incremental commit id. > > Is it possible to add the commit count to the job-name like > sys/conf/newvers.sh in base? > git_cnt=$($git_cmd rev-list --first-parent --count HEAD 2>/dev/null) Places like: https://artifact.ci.freebsd.org/snapshot/stable-13/ have the same sort of problem, partially because the dates(/times) shown at the displayed level need not match what the files will have for date/time (when the files even exist). An example from my past was: https://artifact.ci.freebsd.org/snapshot/stable-13/?C=M&O=D lists ac845558f7b626d9a31b8f6dab686c45d39dc5a0/ as having date/time 2021-Apr-10 18:43 . QUOTE https://artifact.ci.freebsd.org/snapshot/stable-13/ac845558f7b626d9a31b8f6dab686c45d39dc5a0/?C=M&O=D lists: powerpc/ and arm/ as having date/times 2021-Apr-10 18:54 and 2021-Apr-10 18:50 yet lists... i386/ and arm64/ as having date/times 2021-Feb-19 19:00 and 2021-Feb-19 18:50 . But it gets worse: https://artifact.ci.freebsd.org/snapshot/stable-13/ac845558f7b626d9a31b8f6dab686c45d39dc5a0/powerpc/?C=M&O=D shows an empty directory. Same for: https://artifact.ci.freebsd.org/snapshot/stable-13/ac845558f7b626d9a31b8f6dab686c45d39dc5a0/arm/?C=M&O=D By contrast, https://artifact.ci.freebsd.org/snapshot/stable-13/ac845558f7b626d9a31b8f6dab686c45d39dc5a0/i386/?C=M&O=D shows i386/ with date/time 2021-Apr-10 18:43 but https://artifact.ci.freebsd.org/snapshot/stable-13/ac845558f7b626d9a31b8f6dab686c45d39dc5a0/i386/i386/?C=M&O=D shows all the file dates as 2021-Feb-19 19:00 . . . . END QUOTE https://artifact.ci.freebsd.org/snapshot/*/ has already accumulated a huge number of hash-only based naming subdirectories and trying to do a "no-rebuild approximate bisect" via artifact builds has become much more painful/impractical. (It has never helped that architecture is in a deeper layer but is also partial: only some architectures might build for the commit in question. One has had to inspect to find examples of the architecture of interest.) === Mark Millard marklmi at yahoo.com ( dsl-only.net went away in early 2018-Mar) ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: pkg builders order of jobs
On Sun, May 9, 2021 at 3:40 AM Stefan Esser wrote: > Am 09.05.21 um 12:22 schrieb Ronald Klop: > > Hi, > > > > On the pkg builders the list of builds is ordered on git hash. > > http://beefy18.nyi.freebsd.org/jail.html?mastername=main-amd64-default > > (NB: ipv6 only) > > > > This ordering is totally random to me. With svn the list was naturally > ordered > > because of the incremental commit id. > > > > Is it possible to add the commit count to the job-name like > sys/conf/newvers.sh > > in base? > > git_cnt=$($git_cmd rev-list --first-parent --count HEAD 2>/dev/null) > > Adding the start-time of the build in ISO format would also be highly > appreciated, at least by me ;-) > > The date should reflect the time of the last commit that has been > considered. Such a time stamp would let me see with little effort > whether a failure is due to a long running build job that did not > catch-up with the latest fixes to a port, or whether there still is > an issue with a port (especially for the "exotic" architectures that > I do not test on my system and that often need a lot longer to complete > a build job). > > Regards, STefan > I would suggest doing some optimization. Too often a very time consuming build is started well into the process. I mean things that take over 24 hours to build like chromium or electron11. Otherwise on smaller builds, a lot of time can be wasted as one or two slots on the build machines finishing one of the big builds and the other 15 slots idle. This does not happen frequently, but it does happen. With many build cycles exceeding 90 hours, any optimization would be helpful. -- Kevin Oberman, Part time kid herder and retired Network Engineer E-mail: rkober...@gmail.com PGP Fingerprint: D03FB98AFA78E3B78C1694B318AB39EF1B055683 ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: pkg builders order of jobs
Am 09.05.21 um 12:22 schrieb Ronald Klop: > Hi, > > On the pkg builders the list of builds is ordered on git hash. > http://beefy18.nyi.freebsd.org/jail.html?mastername=main-amd64-default > (NB: ipv6 only) > > This ordering is totally random to me. With svn the list was naturally ordered > because of the incremental commit id. > > Is it possible to add the commit count to the job-name like > sys/conf/newvers.sh > in base? > git_cnt=$($git_cmd rev-list --first-parent --count HEAD 2>/dev/null) Adding the start-time of the build in ISO format would also be highly appreciated, at least by me ;-) The date should reflect the time of the last commit that has been considered. Such a time stamp would let me see with little effort whether a failure is due to a long running build job that did not catch-up with the latest fixes to a port, or whether there still is an issue with a port (especially for the "exotic" architectures that I do not test on my system and that often need a lot longer to complete a build job). Regards, STefan OpenPGP_signature Description: OpenPGP digital signature
