atomic script: prepare for f27, multi-arch

dustymabe commented

2017-11-14 06:46:25 +00:00

Member

Quite a few changes in the push_two_week script. Mainly:

enable multi-arch
prepare for f27
ignore autocloud test results for now

Please review each commit individually as that will be easiest for the reviewer.

Quite a few changes in the push_two_week script. Mainly: - enable multi-arch - prepare for f27 - ignore autocloud test results for now Please review each commit individually as that will be easiest for the reviewer.

dustymabe commented

2017-11-14 06:50:26 +00:00

Author

Member

rebased onto e906f1008472ef8560141ea1d69a792e989904f6

dustymabe commented

2017-11-14 06:51:57 +00:00

Author

Member

rebased onto 2848a0a7f6

rebased onto 2848a0a7f6634ee46d5c210dacd1d1ac56c9bcf9

dustymabe commented

2017-11-14 07:41:16 +00:00

Author

Member

14 new commits added

atomic: add multi-arch info to release email
atomic: add dusty to authors of script
atomic: set f26 major release commit for x86_64
atomic: add some hyperlink comments to datagrepper
atomic: rewrite the ostree manipulations code
atomic: rewrite checks, enable multi-arch
atomic: update global vars for multi-arch support
atomic: rename move_ostree_commit -> update_ref
atomic: add get_ostree_compose_info function
atomic: add --ostree-pungi-compose-id option
atomic: add --pungi-compose-id arg, ignore autocloud
atomic: make the parser check for required args
atomic: remove update_ostree_summary_file func
atomic: remove unused release arg

**14 new commits added** * atomic: add multi-arch info to release email * atomic: add dusty to authors of script * atomic: set f26 major release commit for x86_64 * atomic: add some hyperlink comments to datagrepper * atomic: rewrite the ostree manipulations code * atomic: rewrite checks, enable multi-arch * atomic: update global vars for multi-arch support * atomic: rename move_ostree_commit -> update_ref * atomic: add get_ostree_compose_info function * atomic: add --ostree-pungi-compose-id option * atomic: add --pungi-compose-id arg, ignore autocloud * atomic: make the parser check for required args * atomic: remove update_ostree_summary_file func * atomic: remove unused release arg

puiterwijk commented

2017-11-14 13:20:45 +00:00

Contributor

This release branch should no longer be created in the /mnt/koji/compose/updates/atomic/ repo, but rather directly into the /mnt/koji/atomic/$release repo, which means that you can also forget about the sync script for the purposes of this script.

I also can't stop noticing you didn't take my suggestion that the Pungi compose (whose compose ID you pass in the CLI) writes out the exact commitid's, so that you wouldn't have to request this information from datagrepper.

This release branch should no longer be created in the `/mnt/koji/compose/updates/atomic/` repo, but rather directly into the `/mnt/koji/atomic/$release` repo, which means that you can also forget about the sync script for the purposes of this script. I also can't stop noticing you didn't take my suggestion that the Pungi compose (whose compose ID you pass in the CLI) writes out the exact commitid's, so that you wouldn't have to request this information from datagrepper.

mohanboddu commented

2017-11-14 14:26:01 +00:00

Contributor

I think this should be:
update_ref(ostree_commit_data[arch]['ref'], old_commit, ostree_commit_data[arch]['commit'])

Well, old_commit is not actually required, but you are doing some checks to actually move the ref or not in update_ref function. So, its better to update this one.

I think this should be: update_ref(ostree_commit_data[arch]['ref'], old_commit, ostree_commit_data[arch]['commit']) Well, old_commit is not actually required, but you are doing some checks to actually move the ref or not in update_ref function. So, its better to update this one.

dustymabe commented

2017-11-14 14:36:59 +00:00

Author

Member

thanks. I am indeed not passing the right number of args

dustymabe commented

2017-11-14 14:41:45 +00:00

Author

Member

1 new commit added

atomic: operate directly on prod ostree repo

**1 new commit added** * atomic: operate directly on prod ostree repo

dustymabe commented

2017-11-14 14:42:22 +00:00

Author

Member

added new commit that changes to use /mnt/koji/atomic/$release instead

added new commit that changes to use `/mnt/koji/atomic/$release` instead

dustymabe commented

2017-11-14 14:54:06 +00:00

Author

Member

15 new commits added

atomic: operate directly on prod ostree repo
atomic: add multi-arch info to release email
atomic: add dusty to authors of script
atomic: set f26 major release commit for x86_64
atomic: add some hyperlink comments to datagrepper
atomic: rewrite the ostree manipulations code
atomic: rewrite checks, enable multi-arch
atomic: update global vars for multi-arch support
atomic: rename move_ostree_commit -> update_ref
atomic: add get_ostree_compose_info function
atomic: add --ostree-pungi-compose-id option
atomic: add --pungi-compose-id arg, ignore autocloud
atomic: make the parser check for required args
atomic: remove update_ostree_summary_file func
atomic: remove unused release arg

**15 new commits added** * atomic: operate directly on prod ostree repo * atomic: add multi-arch info to release email * atomic: add dusty to authors of script * atomic: set f26 major release commit for x86_64 * atomic: add some hyperlink comments to datagrepper * atomic: rewrite the ostree manipulations code * atomic: rewrite checks, enable multi-arch * atomic: update global vars for multi-arch support * atomic: rename move_ostree_commit -> update_ref * atomic: add get_ostree_compose_info function * atomic: add --ostree-pungi-compose-id option * atomic: add --pungi-compose-id arg, ignore autocloud * atomic: make the parser check for required args * atomic: remove update_ostree_summary_file func * atomic: remove unused release arg

dustymabe commented

2017-11-14 14:54:27 +00:00

Author

Member

amended that last commit to fix typo

dustymabe commented

2017-11-14 15:57:45 +00:00

Author

Member

from @puiterwijk
I also can't stop noticing you didn't take my suggestion that the Pungi compose (whose compose ID you pass in the CLI) writes out the exact commitid's, so that you wouldn't have to request this information from datagrepper.

https://kojipkgs.fedoraproject.org/compose/updates/Fedora-27-updates-20171111.1/logs/x86_64/Everything/ostree-3/commitid.log

There are two reasons why I would prefer to not use a hardcoded URL (like above) for now:

The full path depends on what type of pungi run created the ostree compose. For example the one that we are going to be using for the release today would be:

https://kojipkgs.fedoraproject.org/compose/branched/Fedora-27-20171110.n.1/logs/x86_64/Atomic/ostree-3/commitid.log

I am not really confident in the ostree-N/commitid.log numbering scheme being consistent. For that single compose we have:

logs/aarch64/Everything/ostree-1
logs/ppc64le/Everything/ostree-2
logs/x86_64/Everything/ostree-3

Also, I think when we start creating the workstation ostrees during the updates there will be more ostree-N values in that directory (for x86_64) and we won't know which one to use.

> from @puiterwijk > I also can't stop noticing you didn't take my suggestion that the Pungi compose (whose compose ID you pass in the CLI) writes out the exact commitid's, so that you wouldn't have to request this information from datagrepper. ``` https://kojipkgs.fedoraproject.org/compose/updates/Fedora-27-updates-20171111.1/logs/x86_64/Everything/ostree-3/commitid.log ``` There are two reasons why I would prefer to not use a hardcoded URL (like above) for now: - The full path depends on what type of pungi run created the ostree compose. For example the one that we are going to be using for the release today would be: ``` https://kojipkgs.fedoraproject.org/compose/branched/Fedora-27-20171110.n.1/logs/x86_64/Atomic/ostree-3/commitid.log ``` - I am not really confident in the `ostree-N/commitid.log` numbering scheme being consistent. For that single compose we have: ``` logs/aarch64/Everything/ostree-1 logs/ppc64le/Everything/ostree-2 logs/x86_64/Everything/ostree-3 ``` Also, I think when we start creating the workstation ostrees during the updates there will be more `ostree-N` values in that directory (for x86_64) and we won't know which one to use.

dustymabe commented

2017-11-14 16:13:33 +00:00

Author

Member

15 new commits added

atomic: operate directly on prod ostree repo
atomic: add multi-arch info to release email
atomic: add dusty to authors of script
atomic: set f26 major release commit for x86_64
atomic: add some hyperlink comments to datagrepper
atomic: rewrite the ostree manipulations code
atomic: rewrite checks, enable multi-arch
atomic: update global vars for multi-arch support
atomic: rename move_ostree_commit -> update_ref
atomic: add get_ostree_compose_info function
atomic: add --ostree-pungi-compose-id option
atomic: add --pungi-compose-id arg, ignore autocloud
atomic: make the parser check for required args
atomic: remove update_ostree_summary_file func
atomic: remove unused release arg

**15 new commits added** * atomic: operate directly on prod ostree repo * atomic: add multi-arch info to release email * atomic: add dusty to authors of script * atomic: set f26 major release commit for x86_64 * atomic: add some hyperlink comments to datagrepper * atomic: rewrite the ostree manipulations code * atomic: rewrite checks, enable multi-arch * atomic: update global vars for multi-arch support * atomic: rename move_ostree_commit -> update_ref * atomic: add get_ostree_compose_info function * atomic: add --ostree-pungi-compose-id option * atomic: add --pungi-compose-id arg, ignore autocloud * atomic: make the parser check for required args * atomic: remove update_ostree_summary_file func * atomic: remove unused release arg

dustymabe commented

2017-11-14 16:14:23 +00:00

Author

Member

did a fixup for atomic: add multi-arch info to release email so that the links are sorted (makes for better viewing)

did a fixup for `atomic: add multi-arch info to release email` so that the links are sorted (makes for better viewing)

mohanboddu commented

2017-11-14 16:53:09 +00:00

Contributor

LGTM

@puiterwijk if you can give a quick review(as you know the process better), then I will merge it.

LGTM @puiterwijk if you can give a quick review(as you know the process better), then I will merge it.

puiterwijk commented

2017-11-14 17:00:26 +00:00

Contributor

I don't like the ostree commitid lookup, but let's just get this over with.
If @mohanboddu is okay, let's just do this.

I don't like the ostree commitid lookup, but let's just get this over with. If @mohanboddu is okay, let's just do this.

puiterwijk commented

2017-11-14 17:00:43 +00:00

Contributor

Pull-Request has been merged by puiterwijk

walters commented

2017-11-14 22:12:36 +00:00

Contributor

I scanned this and it looks pretty good to me. Nice split commits! On the multi-arch thing, it feels like it'd make our life easier if pungi dropped that data in the compose json or so?

Rows
Columns

atomic script: prepare for f27, multi-arch #7162