v20241216-232848

Fix a bug when swapfile doesn't exist (#6068)

I made a mistake in https://github.com/pytorch/test-infra/pull/6066 and
should have checked if the file exists first.

v20241216-213021

Clean up existing swapfile before trying to allocate a new one (#6066)

Fix a potential bug from
https://github.com/pytorch/test-infra/pull/6058, here is what I see in
the CI when it happens
https://github.com/pytorch/pytorch/actions/runs/12360769784/job/34497379880?pr=143316#step:5:165

v20241216-184119

mandate IMDSv2 on all new ec2 runners (#6061)

This follows the removal of all IMDSv1 calls weeks ago, as well as the
upgrade of sccache on all repos that use it.

test plan: 

We'll monitor impact, but current IMDSv1 calls are close to 0, with
those left being ones based of a main older than a few weeks.

mitigation in case of failure:
- revert this
-
**https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/configuring-IMDS-existing-instances.html#modify-restore-IMDSv1**
for existing runners

v20241213-212009

Adds swap space for ALI runners and report to ci and metrics if it is…

v20241212-001105

Bugfix: Overprovision scaleups on ephemeral instances & extend runner…

v20241212-000204

Bugfix: Overprovision scaleups on ephemeral instances & extend runner…

v20241114-013723

Revert #5397 now that AL2023/Node 20 rollout is done (#5916)

This is the fallback we have been using for a while to avoid the
upgrade.

v20241112-211933

set hop limit to 3 for IMDSv2 PUT calls for Windows runners (#5897)

v20241111-181517

change the hoplimit for IMDSv2 to 3, to allow docker builds to fetch …

v20241111-180704

bump to support tflint 0.54 and pin version (#5895)