Skip to content

allow_watch_bookmarks does not work #1136

Closed

Description

@NikPaushkin

What happened (please include outputs or screenshots):
Kubernetes API response:

Response is not chunked. Header 'transfer-encoding: chunked' is missing.

What you expected to happen:
Python client should add transfer-encoding header, when allow_watch_bookmarks is True https://kubernetes.io/docs/reference/using-api/api-concepts/#watch-bookmarks

How to reproduce it (as minimally and precisely as possible):
Run watcher for any object list.

Anything else we need to know?:

Environment:

  • Kubernetes version (kubectl version): 1.17.0
  • OS (e.g., MacOS 10.13.6): Windows 10
  • Python version (python --version): 3.8
  • Python client version (pip list | grep kubernetes): 11.0.0

Activity

  1. roycaihw commented on Apr 13, 2020

    @roycaihw
    Member

    /assign @yliaog

  2. fejta-bot commented on Jul 12, 2020

    @fejta-bot

    Issues go stale after 90d of inactivity.
    Mark the issue as fresh with /remove-lifecycle stale.
    Stale issues rot after an additional 30d of inactivity and eventually close.

    If this issue is safe to close now please do so with /close.

    Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
    /lifecycle stale

  3. added
    lifecycle/staleDenotes an issue or PR has remained open with no activity and has become stale.
    on Jul 12, 2020
  4. NikPaushkin commented on Jul 13, 2020

    @NikPaushkin
    Author

    /remove-lifecycle-state

  5. yliaog commented on Jul 13, 2020

    @yliaog
    Contributor

    transfer-encoding header is set at the server side, not at the client side, so it should not be added to python client library

  6. NikPaushkin commented on Jul 20, 2020

    @NikPaushkin
    Author

    @yliaog Thank you for reply. We have this header enabled on servers by default. These are headers from API request:

    content-type: application/json
    transfer-encoding: chunked
    date: Mon, 20 Jul 2020 12:42:03 GMT
    

    Maybe you know what the problem might be?

  7. yliaog commented on Jul 21, 2020

    @yliaog
    Contributor

    it is not obvious to me what might be the problem.
    what if you 'GET' or 'LIST' a resource? if there is no problem there, what about 'WATCH' without allow_watch_bookmarks?

  8. NikPaushkin commented on Jul 21, 2020

    @NikPaushkin
    Author

    @yliaog All work as usual, because as I understand nothing requires 'transfer-encoding: chunked' except of allow_watch_bookmarks

  9. yliaog commented on Jul 21, 2020

    @yliaog
    Contributor

    https://github.com/kubernetes/kubernetes/search?q=transfer-encoding%3A+chunked&unscoped_q=transfer-encoding%3A+chunked

    shows code with 'transfer-encoding: chunked'.

    in watch.go, it was set at the following line.
    https://github.com/kubernetes/kubernetes/blob/34f349257f4fc14c1adc8d9a62109216a6955035/staging/src/k8s.io/apiserver/pkg/endpoints/handlers/watch.go#L201

    I'm not sure since when it was set, could you please check your server version? is it a recent version that has that line?

  10. NikPaushkin commented on Jul 21, 2020

    @NikPaushkin
    Author

    @yliaog Yes, I saw this line as well, have the same version. That's why I have posted an issue here: I was pretty sure that the problem is on the client side.

  11. fejta-bot commented on Aug 20, 2020

    @fejta-bot

    Stale issues rot after 30d of inactivity.
    Mark the issue as fresh with /remove-lifecycle rotten.
    Rotten issues close after an additional 30d of inactivity.

    If this issue is safe to close now please do so with /close.

    Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
    /lifecycle rotten

  12. added
    lifecycle/rottenDenotes an issue or PR that has aged beyond stale and will be auto-closed.
    and removed
    lifecycle/staleDenotes an issue or PR has remained open with no activity and has become stale.
    on Aug 20, 2020
  13. fejta-bot commented on Sep 19, 2020

    @fejta-bot

    Rotten issues close after 30d of inactivity.
    Reopen the issue with /reopen.
    Mark the issue as fresh with /remove-lifecycle rotten.

    Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
    /close

  14. k8s-ci-robot commented on Sep 19, 2020

    @k8s-ci-robot
    Contributor

    @fejta-bot: Closing this issue.

    Details

    In response to this:

    Rotten issues close after 30d of inactivity.
    Reopen the issue with /reopen.
    Mark the issue as fresh with /remove-lifecycle rotten.

    Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
    /close

    Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

Labels

kind/bugCategorizes issue or PR as related to a bug.lifecycle/rottenDenotes an issue or PR that has aged beyond stale and will be auto-closed.

Type

No type

Projects

No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions