Secrets not removed if pod fails admission #144
Description
If a pod fails admission for some reason (e.g. another webhook, mutating or validating, fails) after the telegraf-operator has processed the pod then the resulting secret is not automatically removed.
Unless you're explicitly monitoring and alarming on the number of secret objects in the cluster this can quckly become a problem, for example if a rouge deployment attempting to constantly scale but failing on ValidatingAdmissionWebhooks, eventually causing API server issues if the number of secret objects grows excessively large.
What products and version are you using?
Kubernetes: 1.28
Telegraf-operator: 1.3.11
Additional Comments
It appears this project is no longer being updated or maintained, which is unfortunate as it solves an inherent problem with using influxdb and telegraf to monitor applications in Kubernetes.
As such I decided to completely re-write the project from scratch, resolving this issue, and others such as #137. The project can be found at https://github.com/jmickey/telegraf-sidecar-operator.