This project is mirrored from https://gitee.com/rancher/rancher.git.
Pull mirroring failed .
Repository mirroring has been paused due to too many failed attempts. It can be resumed by a project maintainer.
Repository mirroring has been paused due to too many failed attempts. It can be resumed by a project maintainer.
- 07 Apr, 2020 7 commits
-
-
Daishan Peng authored
Check nil pointer for node.status.NodeConfig
-
GGGitBoy authored
-
loganhz authored
-
Darren Shepherd authored
Fix 502 masking steve errors
-
Darren Shepherd authored
-
Darren Shepherd authored
Only set TLS when tls dialer is not nil
-
Darren Shepherd authored
-
- 06 Apr, 2020 2 commits
- 04 Apr, 2020 1 commit
-
-
Bogdan Girman authored
-
- 03 Apr, 2020 19 commits
-
-
soumyalj authored
Parametrized relevant rbac tests
-
Kinara Shah authored
share-mnt agent shouldn't send back node version
-
Darren Shepherd authored
Lock dynamic TLS to server-url value
-
kinarashah authored
Problem: If there is an extra share-mnt unmanaged agent running on node, it doesn't execute plan but sends back node version and server incorrectly assumes node's been upgraded Solution: ExecutePlan doesn't update processes if writeCertOnly, so add check for it and don't consider it an upgrade
-
Darren Shepherd authored
Support ctx in dialers
-
sowmyav27 authored
-
Darren Shepherd authored
-
Darren Shepherd authored
-
Darren Shepherd authored
-
Darren Shepherd authored
-
Darren Shepherd authored
-
Dan Ramich authored
Fix the issue that users cannot remove proxyUrl of notifier
-
GGGitBoy authored
-
rmweir authored
Eks edit other
-
Prachi Damle authored
Remove the member attribute of the group query
-
GGGitBoy authored
ProxyUrl could be empty. It will alwasy set a empty proxyUrl in alert manger for the notifier. So alert manager cannot send out alerts due to the empty proxy url setting. https://github.com/rancher/rancher/issues/26089
-
Denise Schannon authored
-
rmweir authored
-
rmweir authored
-
- 02 Apr, 2020 4 commits
-
-
Dax McDonald authored
Enqueue controller if still in upgrading state
-
Dax McDonald authored
While upgrading k3s clusters the controller may cause updates that do not change the cluster object. This means that the controller will not be run again to exit the updating state. In these cases the controller should re-enqueue itself.
-
Rajashree Mandaogane authored
Change scope to subtree for openldap login search
-
Rajashree Mandaogane authored
Fix flaky tests
-
- 01 Apr, 2020 3 commits
-
-
rajashree authored
-
Daishan authored
-
Darren Shepherd authored
The resource name is takes from the Norman schema PluralName field. In norman this is a plural form of the ID field which is typically camelCase. A Kubernetes resource field is always lowerscase. In Rancher <=2.3 the RBAC system would do case insensitive checks against all strings. In Rancher 2.4 the RBAC system was switch from Norman to Steve. Steve's implementation is much faster and more proper and as a result does not do case insensitive checks. It is expected that all input to the RBAC system is proper. This is the one place in Rancher where we weird not generating a proper resource name.
-
- 31 Mar, 2020 1 commit
-
-
Prachi Damle authored
-
- 30 Mar, 2020 1 commit
-
-
rajashree authored
-
- 29 Mar, 2020 2 commits
-
-
Darren Shepherd authored
Fix manifest build
-
Darren Shepherd authored
-