Introduce configurable timeout & trivy report size for OCS scanning pod

Problem to solve

The timeout for a scanning pod of a namespace is hardcoded to 5 minutes right now. For OCS scans of a large namespace which might require a longer time, the scan fails due to the timeout. We should enable users to configure the timeout of the OCS scan.

Implementation Plan

• 1. Introduce a scanner_timeout field in the OCS agent configuration.
  • Consider using the google.protobuf.Duration type
  • Define a default timeout if this field is not set
• 2. Configure the scanning pod timeout with this value with an additional 5 minutes for reading the chained config map and transmitting the vuln report.
• 3. Configure this timeout as an env var for the scanning pod
• 4. Configure this timeout for the trivy scan in the trivy-k8s-wrapper repo based on the trivy timeout config
• 5. Create a new release of trivy-k8s-wrapper
• 6. Bump the version of trivy-k8s-wrapper used by gitlab agent
• 6.5: Make the Trivy report size configurable
• 7. Update OCS documentation with steps to configure timeout and report file size. Note that for timeout we would only be able to support configuration in seconds.