Bug 1393774 (tcmigration_cleanup)

[tracking] Buildbot to Taskcluster migration general cleanup

RESOLVED FIXED

Status

task
P4
normal
RESOLVED FIXED
2 years ago
Last year

People

(Reporter: aobreja, Unassigned)

Tracking

Details

Attachments

(2 attachments)

No description provided.
Summary: [tracking] Buildbot and AWS cleanup → [tracking] Buildbot to Taskcluster migration general cleanup
Note explaining the priority level: P4 doesn't mean we've lowered the priority but the contrary. However, we're aligning these levels to the quarterly deliverables, where P1-P3 are taken by our daily waterline KTLO operational tasks.
Priority: -- → P4
Alias: tcmigration_cleanup
windows 8 IX machines - Now - down to 40 remaining machines
windows 8 IX machines - November 20th - down to 20 remaining machines

I think the rest of the hardware is good as we still need it for trunk/beta branches.
Andrei is conducting our first iteration on disabling a bunch of machines so assigning this to him for now.
Assignee: nobody → aobreja
As a first step, after investigation, the following machines where disabled:

Linux
-    bld-linux64-spot   disabled [351-400]  (50 machines)
-    try-linux64-spot   disabled [021-050] ; [300-349] (80 machines)
-     tst-linux64-spot  disabled: [051-100]; [300-399] (150 machines)
-     tst-linux32-spot  disabled: [101-150] ; [300-449] (200 machines)
Windows
-     b-2008-spot disabled: [101-200];[994;999]  (101 machines)
-     y-2008-spot disabled: [102-200] (98 machines)
-     g-w732-spot disabled: [051-200] (150 machines)
-     t-w732-spot disabled: [202-599] (398 machines)
-     t-w864-ix disabled:   [063-101]; [103-115]  (52 machines)
OS X
-    bld-lion-r5 -build disabled: [073-086];[088-092];94 (17 machines)
-    bld-lion-r5 -try disabled:  [22-26];[29-30];[32-35]  (11 machines)

We will keep monitoring to see how this impact production and if everything goes well,we will decommission those that where disabled.
andrei, as per the decommission meeting earlier this week, let's proceed with decommissioning the machines we disabled from comment 4, aside from the 15 slaves we are repurposing as xp in  bug 1410024
Flags: needinfo?(aobreja)
Depends on: 1410024
See Also: → 1409439
Depends on: 1410108
Depends on: 1410109
Depends on: 1410110
Jordan so I will not decomm only those machines disabled by me(mentioned in comment4) but all that are now disabled from all the pools, right?
 The patch for BB look something like this and we will keep only the enable machines.
Flags: needinfo?(aobreja) → needinfo?(jlund)
Attachment #8920586 - Flags: review?(jlund)
Patch for Puppet Sysadmin to remove the machines that were disabled(Nagios change).
Attachment #8921063 - Flags: review?(jlund)
(In reply to Andrei Obreja [:aobreja][:buildduty] from comment #6)
> Created attachment 8920586 [details] [diff] [review]
> bug1393774_bb.patch
> 
> Jordan so I will not decomm only those machines disabled by me(mentioned in
> comment4) but all that are now disabled from all the pools, right?
>  The patch for BB look something like this and we will keep only the enable
> machines.

We can disable any machine that has been disabled for at least the same amount of time as comment 4

But let's not disable anymore until after 57 cycle.
Flags: needinfo?(jlund)
Comment on attachment 8920586 [details] [diff] [review]
bug1393774_bb.patch

Review of attachment 8920586 [details] [diff] [review]:
-----------------------------------------------------------------

I didn't sanity check each against slavealloc but code and logic wise, it lgtm. Perhaps we should add or remove some of the previous comments so that it reflects the truth after this patch?

::: mozilla-tests/production_config.py
@@ +42,5 @@
>  # Bug 1302530 - Add ondemand g-w732 instances
>  for i in range(1, 101):
>      SLAVES['win7_vm_gfx']['g-w732-ec2-%03i' % i] = {}
>  
> +for i in range(2, 3) + range(13, 63):  # Omit 102 for win10 // Bug 1191481 // Bug 1255812 // Bug 1299468 // Bug 1317723 //Bug1397225 //Bug1397879 //Bug1398238 //Bug1398884 //Bug 1399910

I guess this patch is applied after https://bug1410024.bmoattachments.org/attachment.cgi?id=8920094 ?
Attachment #8920586 - Flags: review?(jlund) → review+
Attachment #8921063 - Flags: review?(jlund) → review+
I will unassigned this bug from myself since I will no longer be working on Buildduty,as a note for the next owner this bug can be close once the the dependent bugs are resolved (1372229, 1410024).
Assignee: aobreja → nobody
Bulk change of QA Contact to :jlund, per https://bugzilla.mozilla.org/show_bug.cgi?id=1428483
QA Contact: catlee → jlund
Status: NEW → RESOLVED
Closed: Last year
Resolution: --- → FIXED
Product: Release Engineering → Infrastructure & Operations
You need to log in before you can comment on or make changes to this bug.