Tags: LucasRoesler/faas-netes
Tags
Log and run each informer explicitly Since adding a wait for the cache to sync for Kubernetes endpoints, there appears to be conditions where the code waits indefinitely. The issue can be reproduced by scaling the gateway to zero replicas and back to 1, some of the time it passes the sync, other times it gets stuck. This patch appears to fix the problem and has been tested in controller and operator mode on an RPi 4x node cluster and with k3d with hey running a load test into the nodeinfo function with a minimum of 10 replicas. Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alexellis2@gmail.com>
Set allowPrivilegeEscalation to false as a default This commit fixes openfaas#523 and sets a default, which nobody has raised any concerns with. If this affects you in a negative way then please feel free to raise an issue for discussion. Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alexellis2@gmail.com>
Sync endpoints before starting HTTP server The HTTP server which is used for CRUD and invocations should not be started until the cached informers for endpoints is ready. This may be related to issue openfaas#749 The cache sync duration is now being logged and measured for users to provide logs and additional information. Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alexellis2@gmail.com>
Enable Go durations for timeouts This affects the operator, which was failing to parse the durations being passed to it as Go durations. It supported ints, but wasn't required since this is code already exists and is used for the controller. Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alexellis2@gmail.com>
Fix issue with filtering deployments Fixes issue: openfaas#736 raised by pkoelbl where all deployments in the openfaas-fn namespace were being returned via the REST API when listing functions. The selector's Add method was generating a new value to be returned instead of updating itself. A negative test would have shown this and may be worth adding to the codebase at a later date, or being mocked/stubbed. A PR would be welcomed for this. Thanks also to @LucasRoesler for helping to investigate and confirm the fix. Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alexellis2@gmail.com>
Handle null env vars There was an error related to the following issue in openfaas-cloud, which will also occur in faas-cli, where null env-vars were serialized and passed into the spec. This change has been tested and fixes the problem. openfaas/openfaas-cloud#707 Signed-off-by: Alex Ellis (OpenFaaS Ltd) <alexellis2@gmail.com>
PreviousNext