Relates to https://github.com/api3dao/tasks/issues/695
We've noticed that we are sending too much logs which is costly. The Signed APIs have been redeployed with level INFO to reduce the costs, but pretty much all of the INFO logs come from the server when the request is received (which is arguably still too often for INFO level). Better information would be to:
- log some stats periodically (e.g. how many requests have been received throughout some period).
- log out debug level information (request body, headers) when there is some validation error.
Similar analysis can be done for Airnode feed, which should be even more reasonable with log usage because its deployed by API providers.
Relates to https://github.com/api3dao/tasks/issues/695
We've noticed that we are sending too much logs which is costly. The Signed APIs have been redeployed with level INFO to reduce the costs, but pretty much all of the INFO logs come from the server when the request is received (which is arguably still too often for INFO level). Better information would be to:
Similar analysis can be done for Airnode feed, which should be even more reasonable with log usage because its deployed by API providers.