Skip to content

Releases: kvij/scuttle

v1.1.5

10 May 13:51
Compare
Choose a tag to compare

What's Changed

Thanks for reporting #30 @vitalii-buchyn-exa!

Full Changelog: v1.1.4...v1.1.5

v1.1.4

01 Apr 09:59
Compare
Choose a tag to compare

v1.1.3

16 Dec 10:16
Compare
Choose a tag to compare

What's Changed

Full Changelog: v1.1.2...v1.1.3

v1.1.2

18 Nov 19:33
Compare
Choose a tag to compare

What's Changed

  • Fix pipeline by @kvij in #20
    • Added ARM64 version
    • Fixed issue where Github release artifacts are not build with intended Go version.
      NOTE: Go version is now managed in .github/workflows/release.yaml
  • Update Go version to 1.21.4 by @kvij in #21

Full Changelog: v1.1.1...v1.1.2

v1.1.1

22 Oct 09:23
Compare
Choose a tag to compare

What's Changed

Full Changelog: v1.1.0...v1.1.1

v1.1.0

01 Oct 21:48
Compare
Choose a tag to compare

Feature release with behavior changes

What's new

  • ISTIO_QUIT_API is configured automatically using ENVOY_ADMIN_API as a base if it is configured with the default port. Setting ISTIO_QUIT_API manually to anything but "" will take precedence.
  • Introduced QUIT_REQUEST_TIMEOUT config variable to time out quit requests to ISTIO_QUIT_API and GENERIC_QUIT_ENDPOINTS
  • Introduced GENERIC_QUIT_ONLY config variable to send a request to the configured GENERIC_QUIT_ENDPOINTS but not to ISTIO_QUIT_API when the wrapped application exits.

Behavior changes

ISTIO_QUIT_API set by default in most configurations
If the ENVOY_ADMIN_API is set to the default port 15000 the ISTIO_QUIT_API configured correctly by default.
On deployments that restart the containers the shutdown behavior might be undesirable.
The envoy proxy will be restarted but when the pod has multiple containers that means that meaningful work in the containers that did not crash could be interrupted.
So for pods that have multiple containers that need traffic through the service mesh, it is recommended to set NEVER_KILL_ISTIO to true.

Default timeout of 5 seconds on quit requests
Request to the /quitquitquit endpoint on the pilot agent and requests to GENERIC_QUIT_ENDPOINTS by default time out after 5 seconds. Change the newly introduced QUIT_REQUEST_TIMEOUT variable to override the 5 second default.

Generic quit endpoints called concurrently
When multiple GENERIC_QUIT_ENDPOINTS are configured the requests are no longer waiting on each other but all send at the same time. This might matter if you rely on a configured order for orchestration.

Switched to stdlib HTTP client
All save one dependencies have been cleaned up but that does mean that some fancy http2 stuff that monzo/typhon used to do has also been removed.

What's Changed

  • Update base image to 1.21.1 by @kvij in #16
  • Remove github.com/monzo/typhon dependency by @kvij in #12
  • Add quit request timeout by @kvij in #18
  • Remove PKILL implementation for Istio versions < 1.3 by @kvij in #17
  • Generic quit only by @kvij in #7

Full Changelog: v1.0.4...v1.1.0

v1.0.4

06 Aug 18:59
Compare
Choose a tag to compare

What's Changed

Full Changelog: v1.0.3...v1.0.4

v1.0.3

12 Jul 20:08
Compare
Choose a tag to compare

What's Changed

New Contributors

v1.0.2

29 May 09:25
Compare
Choose a tag to compare

What's Changed

New Contributors

Full Changelog: v1.0.1...v1.0.2

v1.0.1

22 Apr 12:54
Compare
Choose a tag to compare

What's Changed

Full Changelog: v1.0.0...v1.0.1