Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

NoDelayProvisionStrategy won't provision when starting with 0 instances in auto-scaling fleet #180

Closed
jzila opened this issue Apr 10, 2020 · 10 comments

Comments

@jzila
Copy link

jzila commented Apr 10, 2020

I've found that if I let my fleet drop to 0 instances, and then wait a while, the plugin seems to lose track of what's going on and won't provision instances until I get up to 2 jobs pending. This only seems to happen with the NoDelayProvisionStrategy.

Apr 10, 2020 6:55:15 AM INFO com.amazon.jenkins.ec2fleet.NoDelayProvisionStrategy apply
currentDemand -2 availableCapacity 3 (availableExecutors 0 connectingExecutors 1 plannedCapacitySnapshot 2 additionalPlannedCapacity 0)
...
Apr 10, 2020 6:56:25 AM INFO com.amazon.jenkins.ec2fleet.NoDelayProvisionStrategy apply
currentDemand -1 availableCapacity 2 (availableExecutors 0 connectingExecutors 1 plannedCapacitySnapshot 1 additionalPlannedCapacity 0)
@jzila
Copy link
Author

jzila commented Apr 10, 2020

The issue seems to be with plannedCapacitySnapshot, which just seems to be incorrect.

@jzila
Copy link
Author

jzila commented Apr 10, 2020

Actually it also seems like currentDemand is incorrect.

I have:
2 builds in the queue, 1 executor running a third build, and 1 executor connecting. The logs report:

Apr 10, 2020 7:16:25 AM INFO com.amazon.jenkins.ec2fleet.NoDelayProvisionStrategy apply
currentDemand -1 availableCapacity 3 (availableExecutors 0 connectingExecutors 1 plannedCapacitySnapshot 2 additionalPlannedCapacity 0)

@jzila jzila changed the title NoDelayProvisioningStrategy won't provision when starting with 0 instances in auto-scaling fleet NoDelayProvisionStrategy won't provision when starting with 0 instances in auto-scaling fleet Apr 10, 2020
@terma
Copy link

terma commented Apr 14, 2020

thanks for logs, I spent a lot of time but still can't reproduce this beast, is that stable fleet, I mean you created it a lot of time ago and didn't change settings? just go down to zero and fail, is that correct assumption?

@jzila
Copy link
Author

jzila commented Apr 15, 2020

It's been stable for days and it's still giving me this issue.

currentDemand 0 availableCapacity 1 (availableExecutors 0 connectingExecutors 0 plannedCapacitySnapshot 1 additionalPlannedCapacity 0)

image

@jzila
Copy link
Author

jzila commented Apr 15, 2020

Happy to give you any logs you need.

@jzila
Copy link
Author

jzila commented Apr 15, 2020

Actually I've rebooted Jenkins on the newest version and it seems to be picking up builds for new nodes now. Feel free to ignore this issue for now, I'll report back in a few days and close if there's no more repro.

@terma
Copy link

terma commented Apr 16, 2020

Ok, I assume Jenkins on the newest version means Jenkins version not a plugin?

@jzila
Copy link
Author

jzila commented Apr 16, 2020

Ok, I assume Jenkins on the newest version means Jenkins version not a plugin?

Yep, that's correct.

@jzila
Copy link
Author

jzila commented Apr 22, 2020

OK this has been working fine. Closing.

@cccCody
Copy link

cccCody commented Nov 21, 2023

Did you ever figure out a root cause for this, or did it just start working? I think I'm seeing the same thing and filed #425

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants