[BUG] Issues with multi-cluster set up #228
Assignees
Labels
bug
Something isn't working
Comments
|
Hi, is there any update on this to support k8s v1.22? |
kpharasi
pushed a commit
that referenced
this issue
Dec 19, 2024
* Temporary jenkins file to deploy only in qal * Only Build in new jenkins - testing * boilerplate code for traffic config crd * envoy filter creation code * WIP - refactored to separate Traffic Controller and Remote controllers * adding code for decoupled logic * removed commented code, added functionality for remote worker manager * changing workerMgrMap from normal map to sync.Map to handle concurrent access * test checkin * Bootstrap changes * add dependency, rollout & service controllers * add dependency, rollout & service controllers * service entry not needed for admiral-unification * implemented decideEnvoyFilter logic * passing stopCh to processItem to propogate os.exit signal * added new function to get allkeys of a cache map and improved envoy filter creation * Branch name correction in jenkins * Branch name correction in jenkins (#183) * change to us west2 * Ratelimiting filter logic (#184) * Branch name correction in jenkins * change to us west2 * Change default traffic config namespace * Ratelimiting filter logic (#185) * Branch name correction in jenkins * change to us west2 * Change default traffic config namespace * do not proceed with vs geberation in case of rollout * Ratelimiting filter logic (#186) * Branch name correction in jenkins * change to us west2 * Change default traffic config namespace * do not proceed with vs geberation in case of rollout * adding debug statements to identify GetKey() cache panic * reverting back the envoyfilter namespace to admiral-sync * Handle remote cluster unavailability (#191) * Ratelimiting filter logic (#192) * Handle remote cluster unavailability * remove unwanted space before api version * adding new fields to traffic config crd * completed envoy filter operations * updated error handling for processItem function * fixing crd not registered issue * added workload labels i.e. deployment/rollout labels to envoyfilter * updating traffic config CRD to add timeout for edgeservice spec and handling nil cache panic * Ignore updates in passive mode (#197) * Merge heliograph changes (#198) * Ignore updates in passive mode * merge heliograph changes(temporary) * Ratelimiting filter logic (#199) * Ignore updates in passive mode * merge heliograph changes(temporary) * Update heliograph from admiral-deployment, not from jenkins * Temporary hack to override heliograph tag (#200) * Ignore updates in passive mode * merge heliograph changes(temporary) * Update heliograph from admiral-deployment, not from jenkins * temporary hack to override heliograph tag * undo heliograph changes in jenkins * undo heliograph changes in jenkins (#201) * undo heliograph changes in jenkins (#201) * change envoy filter namespace & move vs creation to a different file * change envoy filter namespace & move vs creation to a different file (#203) * undo heliograph changes in jenkins * change envoy filter namespace & move vs creation to a different file * fixing context * vs deletion * logging improvement, deletion (#205) * undo heliograph changes in jenkins * change envoy filter namespace & move vs creation to a different file * vs deletion * asseralias to be in lowercase * debugging labels issue (#207) * no need of retries in vs * Deploy label debugging (#209) * debugging labels issue * handling map update * adding code changes to correct envoy filters and logging struct (#210) * Envofilter formatting (#211) * adding code changes to correct envoy filters and logging struct * updating service now image * Envofilter formatting (#212) * adding code changes to correct envoy filters and logging struct * updating service now image * envoyfilter namespace changed from admiral-sync to istio-system * changing envoy filter namespace from admiral-sync to istio-system (#213) * Ns change (#214) * changing envoy filter namespace from admiral-sync to istio-system * fixing envoyfilter issue for name and struct * vs env delete fix * handling traffic config isDisbaled annotation, also handling update and delete (#217) * cname identity map pupulation fix * position gervice entry selector only for admiral * altering traffic config persona code flow in service entry * altering traffic config persona code flow in service entry (#221) * making strings as const in common for resue * Unification test (#222) * altering traffic config persona code flow in service entry * making strings as const in common for resue * adding adaptive concurrency filter * altering traffic config persona code flow in service entry * remove cluster from map upon deletion * debug traces * review comments * null check for virtual service * worker manager pointer corrected * add traffic config object for all clusters the asset is in * envoy filter env crud fix * admiral-unification in qal east * handling empty workloadenvselector lable case (#228) * Delete registry.go_backup Not needed * Delete types.go_backup * Delete virtualservice.go_backup * Delete envoyfilter-operations.go_backup * merge conflict for pull req 229 Co-authored-by: Rajendra Gosavi <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Describe the bug
The multi-cluster set up docs aren't clear and there are at least a couple of issues when users tried the set up
The text was updated successfully, but these errors were encountered: