- bump @kubernetes/client-node to 0.23.2
- bump gaxios to 6.7.1
- fix other dependencies
- only support Node 18 or higher
- bump @kubernetes/client-node to 0.20.0
- bump gaxios to 5.1.3
- bump @kubernetes/client-node to 0.18.1
- bump gaxios to 5.1.0
- bump @kubernetes/client-node to 0.18.0
- bump qs from 6.5.2 to 6.5.3
- upgrade dependencies
- (dependabot) bump jose from 4.8.3 to 4.9.3
- made 'logger' protected
- upgrade dev dependencies
- upgrade dependencies
- upgrade dependencies
- support Node 12 or higher
- switch from axios to gaxios
- add new applyGaxiosKubeConfigAuth() method
- upgrade dependencies
- upgrade dependencies
- removed use of serialize-error
- fix vulnerabilities
- upgrade dependencies
- upgrade @kubernetes/client-node to 0.16.1
- upgrade dependencies
- remove support for v1beta1 CRDs
- upgrade dependencies
- upgrade @kubernetes/client-node to 0.15.1
- upgrade async to 3.2.1
- upgrade axios to 0.21.3
- fix CVE-2021-32803
- update @kubernetes/client-node to 0.15.0
- eliminate js-yaml dependency
- update dependencies
- update @kubernetes/client-node to 0.14.3
- update other dependencies
- update @kubernetes/client-node to 0.14.0
- support both apiextensions.k8s.io/v1 and apiextensions.k8s.io/v1beta1
- updated dependencies
- removed local
ForeverWatch
(obsolete due to latest@kubernetes/client-node
) - updated dependencies
- using a local
ForeverWatch
until newer@kubernetes/client-node
is released - updated dependencies
- fixed usage of 'serialize-error'
- updated dependencies
- use 'serialize-error' to log errors
- replaced direct dependency on 'request' with 'axios'
- added stream-buffers to dependencies
- updated @kubernetes/client-node to fix security issue found in node-forge
- reliability enhancement: end process on watch error now (to force a pod restart)
- upgraded to TypeScript 4.0
- updated devDependencies
- updated dependencies
- updated to latest
lodash
to fix vulnerability
- added an
export
toResourceMetaImpl
- fixed missing await on kubeConfig.applyToRequest()
- update dependencies
- Some small tweaking to make error-handling on watches as robust as possible
- Added two debug logs for restarting a watch
watchResource()
now has an optionalnamespace
parameter- updated dependencies
- updated dependency on @kubernetes/client-node to 0.11.2
- fixed minimist security advisory warning
- graduated to stable
- upgraded some dependencies
- back to using
request-promise-native
- removed dependency on
request-promise-native
- upgraded all dependencies
- tweaked
handleResourceFinalizer()
to always return 'true' if the resource is marked for deletion
- switched to ESLint (TSLint is deprecated)
- some small breaking changes to interface names due to compliance with standard ESLint rules now (
ResourceEvent
,ResourceMeta
andOperatorLogger
)
- added
handleResourceFinalizer()
andsetResourceFinalizers()
to easily implement robust handling of the deletion a resource (using finalizers).
- Removed
watchCustomResource()
as it made little sense to use this in practice due to the required permissions. Just usewatchResource()
. - Small tweaks.
- Removed
watchCustomResource()
fromREADME.md
. - Updated to TypeScript 3.6.3.
- Added
patchResourceStatus()
to patch a status object rather than having to set it completely every time. - Fixed problem with all events running in parallel due to async/await. They are processed consecutively now.
- Initial release.