* all credentials moved to globals
* cassandra fix
* centralized certificates, removed unneded entities
* minor fixes, local-multi-namespace example fixes
* removing unneeded sections in the yaml files
* updates to changelog and multiple namespaces examples
* fixing last couple of services, removed not needed secrets, centralized httpclientconfig.json and ssl.properties
* minor improvements
* changelog reformatted
* fixing startupprobe and changelog
Co-authored-by: Gleb Boushev <4c74356b41@outlook.com>
* WIFI-990: disable hardcoded NodePort for AWS deployments
* add prefix to AP debug ports
* update debug port range in port-forwarding run.sh
Co-authored-by: Eugene Taranov <eugene@opsfleet.com>
EFS Provisioner will be instantiated when it's an AWS deployment.
Please make sure to update the values in the dev-amazon-connectus.yaml override with your
own EFS FileSystem
The same results were obtained after converting Portal into a Statefulset and
using PVC as a VolumeClaimTemplate. Since PVC name is fixed, it becomes easy to
get the PVC name and bind it to the other pod (opensync-gw).
The purpose of this Repo is to initialize artifacts that would be used by the
tip-wlan release later on.
For now, we are initializing Namespace and PVC (that's to be shared between 2 deployments).
Besides, we are also exposing ports 9096 and 9097 on Opensync-gw-cloud service. These will
then be linked to the ALB for any traffic coming from APs.
We are not able to generate the mqttData.log if we use IngressController as NodePorts.
Looked at AP, it was generating the metrics but somehow they were not repoted in the mqttData.log file in the GW-sync pod.
Changing the MQTT and GW-SYNC to NodePort service and then changing the LoadBalancers to use TargetGroup relative to these NodePorts
fixed the issue. Still not clear on the reason.
- This also includes Liveliness and Readiness checks for the services(TW-341, TW-345)
- Also added Env variables so they can communicate with each other(TW-342, TW-346)
