How can performance of MetaDefender Core be increased when deployed on K8S environments?

Background

While resource allocation influence the performance of MetaDefender Core, there are also internal environmental variables specific to the MetaDefender Core application that can be updated to increase performance on a Kubernetes deployment as well as we do on regular package installs.

Steps

For Kubernetes deployments, you can make updates to performance parameters using the CORE_CONG_JSON variable locates in the values.yaml file, like is done in metadefender-k8s/helm_charts/mdcore/values.yaml at f13e458b58b730d0735d2624abc4aef655c4487e · OPSWAT/metadefender-k8s

Using this environment variable, different values detailed in MetaDefender Configuration can be set. In particular the ones related to performance like parallecount, db_connection, etc. Keep in mind that the format for these entries should follow the format below:

[Section]/[parameter]: [value]

Example

For example, to set the parallel_count from its default, you can add an entry like: "internal/parallelcount": 40. This line would be updated as follows:

CORE_CONF_JSON: '{"global/restaddress": "0.0.0.0", "internal/parallelcount": 40, "internal/next_extraction_polling_interval": 1000}'