Giters

ncipollo / release-action

An action which manages a github release

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Error 500: Request body length does not match content-length header

dhwz opened this issue · comments

commented

After your action was updated today to v1.14.0 it always fails with the above error, after switching back to v1.13.0 it's working again.

If you search on Google for that error message it seems to be related to the change to Node20.

commented

Could you provide more info on how to reproduce? Do you have a link to a repo which is experiencing this issue?

commented

Yes of course, this is the workflow which is executed on a regular base and it failed since this morning.
https://github.com/AmberELEC/AmberELEC/actions/runs/7795762806/workflow
It failed while running release-prerelease

commented

Thanks! A few more follow up questions:

  1. Is main a custom github runner? (i.e runs on main)? If so, could you check that node is up to date?
  2. Did you ever rename the AmberELEC-prerelease repo? From googling around it seems like there could be some issues with redirects, node 20 and octokit.

I've been testing this in my test repo and have so far been unable to reproduce (https://github.com/ncipollo/actions-playground)

I've tried pretty much everything I can see from your workflow which is slightly out of my typical test path, including:

  • Using a long form body input (as opposed to body file)
  • Publishing to another repo
  • Tried on windows & ubuntu instances

Anything else you can think of I should try?

commented

Also I'm not seeing a job failure in that repo you linked. Do you have a specific job you can link me to, I want to see if there are any logs.

commented

1.) Yes it's our own server for the runner, I've to check what node version is on the runner, I've no access have to ask our admin.
2.) Indeed the repo was renamed previously but the URL above is the current name so it shouldn't have any redirects

This error is strange, now I tried to reproduce it to get you some logs but now it doesn't happen. It failed multiple times this morning.
I'll close the issue and reopen it when I have new information on this topic and also have a failed job.

commented

Hmm quite weird. Thanks for the update!

commented

Were you able to check with your admin about what node version you are using? I wasn't able to reproduce, so I don't think there is much I can do until we narrow down what the root cause is.