The http logging middleware splits out different request results and logs them as either debug or warn. Generally errors are logged as warn and successes are logged as debug.
|
if 100 <= statusCode && statusCode < 500 || statusCode == http.StatusBadGateway || statusCode == http.StatusServiceUnavailable { |
|
l.logWithRequest(r).Debugf("%s %s (%d) %s", r.Method, uri, statusCode, time.Since(begin)) |
|
if l.LogRequestHeaders && headers != nil { |
|
l.logWithRequest(r).Debugf("ws: %v; %s", IsWSHandshakeRequest(r), string(headers)) |
|
} |
|
} else { |
|
l.logWithRequest(r).Warnf("%s %s (%d) %s Response: %q ws: %v; %s", |
|
r.Method, uri, statusCode, time.Since(begin), buf.Bytes(), IsWSHandshakeRequest(r), headers) |
|
} |
We need to log the below error conditions that are currently being logged as debug. Unfortunately, due to volume, we can't turn on debug logging.
statusCode == http.StatusBadGateway || statusCode == http.StatusServiceUnavailable
My guess is that these two statuses are logged at a debug level due to volume if the backend is unavailable. We would like to log these failures at a higher level than debug, but also recognize that the volume would be too great to log if a backend is down.
The change we'd like to make:
- Move
http.StatusBadGateway and http.StatusServiceUnavailable to be logged at a Warn level with the other errors
- Use a configurable rate limited logger to log errors instead of logging 100% of all errors at Warn
Thoughts?
If this (or something similar) is acceptable I'd be glad to PR it.
@bboreham
The http logging middleware splits out different request results and logs them as either debug or warn. Generally errors are logged as warn and successes are logged as debug.
common/middleware/logging.go
Lines 56 to 64 in 4b18475
We need to log the below error conditions that are currently being logged as debug. Unfortunately, due to volume, we can't turn on debug logging.
My guess is that these two statuses are logged at a debug level due to volume if the backend is unavailable. We would like to log these failures at a higher level than debug, but also recognize that the volume would be too great to log if a backend is down.
The change we'd like to make:
http.StatusBadGatewayandhttp.StatusServiceUnavailableto be logged at a Warn level with the other errorsThoughts?
If this (or something similar) is acceptable I'd be glad to PR it.
@bboreham