Additional Nautobots¶
In certain situations Nautobot experiences a heavy load due to high API or GraphQL usage. In these cases it may be desirable to scale out Nautobot specifically for the /api or /api/graphql endpoints to be able to handle this load better without scaling out the default UI deployments. This can be accomplished using the nautobots key in the helm chart.
This will create a separate deployment dedicated to /api requests. Additionally if you wanted to further tweak this specific deployment you can use any of the keys from nautobot which are the default values used by this api Nautobot deployment and override them. For example to increase the memory requests specifically for this deployment to 32G you could set the following:
nautobots:
api:
enabled: true
ingressPaths:
- "/api"
initNautobot: false
resources:
requests:
memory: "32G"
This is completely dynamic so you can create /api/graphql or /plugins deployments depending on your use case.