Maemo repackages of individual locales should be done in a meaningful location

RESOLVED FIXED

Status

Release Engineering
General
P2
normal
RESOLVED FIXED
8 years ago
4 years ago

People

(Reporter: armenzg, Assigned: armenzg)

Tracking

Firefox Tracking Flags

(Not tracked)

Details

Attachments

(1 attachment, 2 obsolete attachments)

(Assignee)

Description

8 years ago
In bug 519684 I am fixing where the Maemo builds are happening under scratchbox to avoid a collision that was happening for the "l10n" directory since it was shared among different branches.

In this bug I am just asking to check out things in a more meaningful way even though there is not collision.

The changes required would be in mobile_config.py and they would be like: 
  ...['linux-gnueabi-arm']['base_builddir'] = 'maemo-trunk'
- ...['linux-gnueabi-arm']['base_l10n_workdir'] = '%s/l10n' % SBOX_HOME
+ ...['linux-gnueabi-arm']['base_l10n_workdir'] = '%s/build/maemo-trunk-l10n' % SBOX_HOME
or something like that where the directory structure would be something like this:

build/
  maemo-trunk
  maemo-trunk-l10n
  maemo-1.9.2
  maemo-1.9.2-l10n
That shouldn't be a huge change, but we will need to make sure that those locations aren't hardcoded in other parts of the code.
Component: Release Engineering → Release Engineering: Future

Comment 2

8 years ago
Armen asked me to take this.

John (Ford): would there be anything in the mobile or l10n repos that would reference this directly, or will I be fine checking buildbot-configs and buildbotcustom?
Assignee: nobody → ccooper
Status: NEW → ASSIGNED
Component: Release Engineering: Future → Release Engineering
OS: Mac OS X → All
Priority: -- → P3
Hardware: x86 → All

Comment 3

8 years ago
(In reply to comment #0)
> In bug 519684 I am fixing where the Maemo builds are happening under scratchbox
> to avoid a collision that was happening for the "l10n" directory since it was
> shared among different branches.
> 
> In this bug I am just asking to check out things in a more meaningful way even
> though there is not collision.
> 
> The changes required would be in mobile_config.py and they would be like: 
>   ...['linux-gnueabi-arm']['base_builddir'] = 'maemo-trunk'
> - ...['linux-gnueabi-arm']['base_l10n_workdir'] = '%s/l10n' % SBOX_HOME
> + ...['linux-gnueabi-arm']['base_l10n_workdir'] = '%s/build/maemo-trunk-l10n' %
> SBOX_HOME
> or something like that where the directory structure would be something like
> this:
> 
> build/
>   maemo-trunk
>   maemo-trunk-l10n
>   maemo-1.9.2
>   maemo-1.9.2-l10n
Priority: P3 → P2

Comment 4

8 years ago
Created attachment 408545 [details] [diff] [review]
Rename mobile l10n workdirs

This is running on staging-master2 right now.
Attachment #408545 - Flags: review?(armenzg)
(Assignee)

Updated

8 years ago
Attachment #408545 - Flags: review?(armenzg) → review+

Comment 5

8 years ago
Created attachment 408677 [details] [diff] [review]
Rename mobile l10n workdirs, production patch
Attachment #408677 - Flags: review?(armenzg)
(Assignee)

Updated

8 years ago
Attachment #408677 - Flags: review?(armenzg) → review+
(Assignee)

Comment 6

8 years ago
Folder "l10n" under scratchbox will not be in used anymore and will require to manually remove it. This can be done with cssh (I have it set up if you want me to do it)

Comment 7

8 years ago
Created attachment 409096 [details] [diff] [review]
Rename mobile l10n workdirs, corrected for bitrot

Carrying forward previous review, just fixing the bitrot.
Attachment #408545 - Attachment is obsolete: true
Attachment #408677 - Attachment is obsolete: true
Attachment #409096 - Flags: review+

Comment 8

8 years ago
Comment on attachment 409096 [details] [diff] [review]
Rename mobile l10n workdirs, corrected for bitrot

http://hg.mozilla.org/build/buildbot-configs/rev/bf93f21951fe
Attachment #409096 - Flags: checked-in+
(Assignee)

Comment 9

8 years ago
I will jump and clean the slaves before the end of today since I have things set up to access all slaves.

I will start doing it no sooner than 40 minutes (repackages take less than 5 minutes) after the reconfiguration of the masters.
production-master* have been reconfig-ed with this change.

Re-assigning to Armen for the slave cleanup.
Assignee: ccooper → armenzg
(Assignee)

Comment 11

8 years ago
I removed SBOX_HOME/l10n from all linux slaves 01 to 49.
I will check again later today to see if everything is OK.
But not slave 50 ?
(Assignee)

Comment 13

8 years ago
(In reply to comment #12)
> But not slave 50 ?
My bad. 50 was done as well.

I went around yesterday and I put the removing processes in the background to be able to don't have to wait for each batch of slaves. When I checked this morning some of it had delete what I wanted and some not.

There is no internet where I am going so I will have to finish this on Monday.
If anyone takes this, all slaves should have a SBOX/l10n.delete folder instead of SBOX/l10n since I moved them before deleting them.

Comment 14

8 years ago
I have nuked /scratchbox/users/cltbld/home/cltbld/build, l10n (if it exists), and l10n.delete on moz2-linux-slave01 through 50... I had to go in there anyway to clobber the Maemo depend build dirs (bug 489410).

This bug looks done; please reopen if I'm resolving prematurely.
Status: ASSIGNED → RESOLVED
Last Resolved: 8 years ago
Resolution: --- → FIXED
(Assignee)

Comment 15

8 years ago
Aki thanks tons!
Product: mozilla.org → Release Engineering
You need to log in before you can comment on or make changes to this bug.