Update the routing logic based on recent changes #9307

hross · 2021-08-25T10:28:43Z

Why:

We have updated the routing logic for runners and want to make it clear in the docs.

What's being changed:

Routing logic for self hosted runners documentation.

Check off the following:

I have reviewed my changes in staging (look for the latest deployment event in your pull request's timeline, then click View deployment).
For content changes, I have completed the self-review checklist.

Writer impact (This section is for GitHub staff members only):

This pull request impacts the contribution experience
- I have added the 'writer impact' label
- I have added a description and/or a video demo of the changes below (e.g. a "before and after video")

welcome · 2021-08-25T10:28:44Z

Thanks for opening this pull request! A GitHub docs team member should be by to give feedback soon. In the meantime, please check out the contributing guidelines.

ramyaparimi · 2021-08-25T13:50:01Z

@hross
Thanks so much for opening a PR! I'll get this triaged for review ⚡

TingluoHuang · 2021-08-25T13:54:27Z

content/actions/hosting-your-own-runners/using-self-hosted-runners-in-a-workflow.md

@@ -74,5 +74,5 @@ When routing a job to a self-hosted runner, {% data variables.product.prodname_d
 2. The job is then sent to the first matching runner that is online and idle.


{% data variables.product.prodname_dotcom %} first searches for an online and enabled runner at the repository level, then at the organization level{% ifversion ghes or ghae %}, then at the enterprise level{% endif %}.

If we don't find an online and enabled runner at any level, the job is queued to all levels and wait for any runner from any level to come online and pickup the job.

If the job remains queued for more than 24 hours, the job will fail.

If we find an online and enabled runner (preferred runner) at a certain level, the job is then sent to the preferred runner.

60 seconds after sending the job, if the job is not picked up by the preferred runner, we will try to send the same job to all other levels as well.

If the job remains queued for more than 24 hours, the job will fail.

Thanks @TingluoHuang, I've updated the draft accordingly ⚡

TingluoHuang · 2021-08-25T14:08:22Z

content/actions/hosting-your-own-runners/using-self-hosted-runners-in-a-workflow.md

@@ -71,8 +71,7 @@ These labels operate cumulatively, so a self-hosted runner’s labels must match
 When routing a job to a self-hosted runner, {% data variables.product.prodname_dotcom %} looks for a runner that matches the job's `runs-on` labels:

 1. {% data variables.product.prodname_dotcom %} first searches for a runner at the repository level, then at the organization level{% ifversion ghes or ghae %}, then at the enterprise level{% endif %}. 
+   - If no online runner is found, the job will be queued to all levels and whichever level first has an online and availabile runner will pick up the job.


If no online runner and enabled is found, the job will be queued to all levels and whichever level first has an online and enabled runner will pick up the job.

Added to draft 👍

martin389 · 2021-08-27T01:02:44Z

Thanks @TingluoHuang -- I've updated the draft with your comments, and this is ready for another review 👍

TingluoHuang · 2021-08-27T14:33:37Z

content/actions/hosting-your-own-runners/using-self-hosted-runners-in-a-workflow.md

   - If the job remains queued for more than 24 hours, the job will fail.
+- If {% data variables.product.prodname_dotcom %} finds an online and enabled runner (preferred runner) at a certain level, the job is then sent to the preferred runner.
+    - If the job is not picked up by the preferred runner within 60 seconds after sending the job, {% data variables.product.prodname_dotcom %} will try send the same job to all other levels as well.


not sure we need to add any more detail after we send the job to all levels.

If the job is not picked up by the preferred runner within 60 seconds after sending the job, {% data variables.product.prodname_dotcom %} will try send the same job to all other levels as well and waits for any runner from any level to come online and pickup the job.

What does it mean that send the same job to all other levels as well.? Is it the same behavior as the earlier description of the job is queued to all levels and waits for any runner from any level to come online and pickup the job.?

Or, is this the same as saying something like: "If the runner doesn't pick up the assigned job within 60 seconds, GitHub starts searching again for an online and enabled runner at all levels."?

We don't search for an online and enabled runner after 60 seconds, we queue the job to all levels and wait for a label matched runner from one of the levels that comes online/enable to pick up the job

lucascosti

I asked some questions and made a suggestion to make this a little clearer.

@TingluoHuang When did these change come in to effect? I assume it won't be included in GHES 3.2?

@martin389 We'll probably need to keep the old description for the GHES versions the new one doesn't apply to yet.

lucascosti · 2021-08-31T05:26:51Z

content/actions/hosting-your-own-runners/using-self-hosted-runners-in-a-workflow.md

   - If the job remains queued for more than 24 hours, the job will fail.
+- If {% data variables.product.prodname_dotcom %} finds an online and enabled runner (preferred runner) at a certain level, the job is then sent to the preferred runner.
+    - If the job is not picked up by the preferred runner within 60 seconds after sending the job, {% data variables.product.prodname_dotcom %} will try send the same job to all other levels as well.


What does it mean that send the same job to all other levels as well.? Is it the same behavior as the earlier description of the job is queued to all levels and waits for any runner from any level to come online and pickup the job.?

Or, is this the same as saying something like: "If the runner doesn't pick up the assigned job within 60 seconds, GitHub starts searching again for an online and enabled runner at all levels."?

lucascosti · 2021-08-31T05:29:30Z

content/actions/hosting-your-own-runners/using-self-hosted-runners-in-a-workflow.md

-   - If all matching runners are offline, the job will queue at the level with the highest number of matching offline runners.
-   - If there are no matching runners at any level, the job will fail.
+- {% data variables.product.prodname_dotcom %} first searches for an online and enabled runner at the repository level, then at the organization level{% ifversion ghes or ghae %}, then at the enterprise level{% endif %}.
+- If {% data variables.product.prodname_dotcom %} doesn't find an online and enabled runner at any level, the job is queued to all levels and waits for any runner from any level to come online and pickup the job.


@TingluoHuang In the previous description, we said that If there are no matching runners at any level, the job will fail.. With this new behavior, if there are no runners configured at any level that match the specified labels for the job, will the job be queued and wait 24 hours before failing?

The job will be queued and wait for 24 hours before failing. Within 24 hours, any label matched runner from any level (repo/org/enterprise) that comes online can pick up the job

content/actions/hosting-your-own-runners/using-self-hosted-runners-in-a-workflow.md

TingluoHuang · 2021-08-31T12:38:53Z

@lucascosti the change is NOT in GHES 3.2

lucascosti · 2021-09-01T05:41:55Z

Ok wording is ready for review:

https://docs-9307--hross-update-assign.herokuapp.com/en/actions/hosting-your-own-runners/using-self-hosted-runners-in-a-workflow#routing-precedence-for-self-hosted-runners

@TingluoHuang / @hross could you please confirm its accuracy?

I've opened a docs-engineering issue internally to look at the check that is failing.

TingluoHuang · 2021-09-01T17:05:16Z

content/actions/hosting-your-own-runners/using-self-hosted-runners-in-a-workflow.md

@@ -70,9 +70,17 @@ These labels operate cumulatively, so a self-hosted runner’s labels must match

 When routing a job to a self-hosted runner, {% data variables.product.prodname_dotcom %} looks for a runner that matches the job's `runs-on` labels:

-1. {% data variables.product.prodname_dotcom %} first searches for a runner at the repository level, then at the organization level{% ifversion ghes or ghae %}, then at the enterprise level{% endif %}. 
+{% ifversion fpt or ghes > 3.2 or ghae %}


this behavior is not on for GHAE M1, not sure whether that matters to the doc.

🤔 Hmm, ok; I'll edit this for -next

TingluoHuang · 2021-09-01T17:07:02Z

content/actions/hosting-your-own-runners/using-self-hosted-runners-in-a-workflow.md

+  - If the runner doesn't pick up the assigned job within 60 seconds, the job is queued at all levels and waits for a matching runner from any level to come online and pick up the job.
+- If {% data variables.product.prodname_dotcom %} doesn't find an online and idle runner at any level, the job is queued to all levels and waits for a matching runner from any level to come online and pick up the job.
+- If the job remains queued for more than 24 hours, the job will fail.
+{% elsif ghes < 3.3 %}


<= 3.2 ? 😆 I saw the linter error.

haha, unfortunately, we can't use <= or >= in our liquid helper 🙁

github-actions · 2021-09-03T07:00:07Z

Thanks very much for contributing! Your pull request has been merged 🎉 You should see your changes appear on the site in approximately 24 hours. If you're looking for your next contribution, check out our help wanted issues ⚡

Update the routing logic based on recent changes

6d161c3

hross requested a review from TingluoHuang August 25, 2021 10:28

github-actions bot added the triage Do not begin working on this issue until triaged by the team label Aug 25, 2021

hross requested a review from martin389 August 25, 2021 10:30

ramyaparimi added content This issue or pull request belongs to the Docs Content team waiting for review Issue/PR is waiting for a writer's review and removed triage Do not begin working on this issue until triaged by the team labels Aug 25, 2021

TingluoHuang reviewed Aug 25, 2021

View reviewed changes

Update using-self-hosted-runners-in-a-workflow.md

52854cd

TingluoHuang reviewed Aug 25, 2021

View reviewed changes

hross and others added 2 commits August 26, 2021 13:40

Update using-self-hosted-runners-in-a-workflow.md

c530a57

Merge branch 'main' into hross-update-assign-logic

0c23442

martin389 self-assigned this Aug 26, 2021

Update using-self-hosted-runners-in-a-workflow.md

bef007e

martin389 requested a review from TingluoHuang August 27, 2021 01:01

Merge branch 'main' into hross-update-assign-logic

ef22a15

TingluoHuang reviewed Aug 27, 2021

View reviewed changes

TingluoHuang previously approved these changes Aug 27, 2021

View reviewed changes

lucascosti reviewed Aug 31, 2021

View reviewed changes

Rewrite conditions; add versioning for previous GHES versions

5aba563

lucascosti dismissed TingluoHuang’s stale review via 5aba563 September 1, 2021 04:52

Merge branch 'main' into hross-update-assign-logic

fd89767

lucascosti self-assigned this Sep 1, 2021

Add versioning to enterprise level mention

b5f740a

TingluoHuang reviewed Sep 1, 2021

View reviewed changes

TingluoHuang previously approved these changes Sep 1, 2021

View reviewed changes

Changed versioning for GHAE and GHES < 3.3

6d1341a

lucascosti dismissed TingluoHuang’s stale review via 6d1341a September 3, 2021 04:29

Merge branch 'main' into hross-update-assign-logic

7291f76

martin389 approved these changes Sep 3, 2021

View reviewed changes

Merge branch 'main' into hross-update-assign-logic

1d912f1

lucascosti enabled auto-merge (squash) September 3, 2021 06:50

lucascosti merged commit 49a9224 into main Sep 3, 2021

lucascosti deleted the hross-update-assign-logic branch September 3, 2021 06:59

0x2b3bfa0 mentioned this pull request Nov 29, 2021

Remove useless runner waiting code iterative/terraform-provider-iterative#315

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Update the routing logic based on recent changes #9307

Update the routing logic based on recent changes #9307

hross commented Aug 25, 2021

welcome bot commented Aug 25, 2021

ramyaparimi commented Aug 25, 2021

TingluoHuang Aug 25, 2021

martin389 Aug 27, 2021

TingluoHuang Aug 25, 2021 •

edited

Loading

martin389 Aug 27, 2021

martin389 commented Aug 27, 2021

TingluoHuang Aug 27, 2021

lucascosti Aug 31, 2021

TingluoHuang Aug 31, 2021

lucascosti left a comment

lucascosti Aug 31, 2021

lucascosti Aug 31, 2021

TingluoHuang Aug 31, 2021

TingluoHuang commented Aug 31, 2021

lucascosti commented Sep 1, 2021

TingluoHuang Sep 1, 2021

lucascosti Sep 3, 2021

TingluoHuang Sep 1, 2021

lucascosti Sep 3, 2021

github-actions bot commented Sep 3, 2021

		@@ -74,5 +74,5 @@ When routing a job to a self-hosted runner, {% data variables.product.prodname_d
		2. The job is then sent to the first matching runner that is online and idle.

Update the routing logic based on recent changes #9307

Update the routing logic based on recent changes #9307

Conversation

hross commented Aug 25, 2021

Why:

What's being changed:

Check off the following:

Writer impact (This section is for GitHub staff members only):

welcome bot commented Aug 25, 2021

ramyaparimi commented Aug 25, 2021

Choose a reason for hiding this comment

Choose a reason for hiding this comment

TingluoHuang Aug 25, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

martin389 commented Aug 27, 2021

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

lucascosti left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

TingluoHuang commented Aug 31, 2021

lucascosti commented Sep 1, 2021

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

github-actions bot commented Sep 3, 2021

TingluoHuang Aug 25, 2021 •

edited

Loading