linkerd2/controller
Andrew Seigner 50f4aa57e5
Require timestamp on all telemetry requests (#342)
PR #298 moved summary (non-timeseries) requests to Prometheus' Query
endpoint, with no timestamp provided. This Query endpoint returns a
single data point with whatever timestamp was provided in the request.
In the absense of a timestamp, it uses current server time. This causes
the Public API to return discreet data points with slightly different
timestamps, which is unexpected behavior.

Modify the Public API -> Telemetry -> Prometheus request path to always
require a timestamp for single data point requests.

Fixes #340

Signed-off-by: Andrew Seigner <siggy@buoyant.io>
2018-02-13 13:52:21 -08:00
..
api Require timestamp on all telemetry requests (#342) 2018-02-13 13:52:21 -08:00
cmd Set conduit version to match conduit docker tags (#208) 2018-01-26 11:43:45 -08:00
destination controller: echo ip address if destination service receives ip (#186) 2018-01-22 16:20:13 -08:00
gen Remove unused metrics (#322) 2018-02-09 17:14:17 -08:00
k8s Better error handling for Tap (#177) 2018-01-25 11:49:38 -05:00
script Remove unused metrics (#322) 2018-02-09 17:14:17 -08:00
tap Better error handling for Tap (#177) 2018-01-25 11:49:38 -05:00
telemetry Require timestamp on all telemetry requests (#342) 2018-02-13 13:52:21 -08:00
util Add more prometheus instrumentation (latency, response size) (#174) 2018-02-01 09:50:31 -08:00
Dockerfile Precompile large Go libraries in go-deps Docker image. (#332) 2018-02-13 11:35:10 -10:00