search

OpenTSDB / opentsdb

A scalable, distributed Time Series Database.
http://opentsdb.net
GNU Lesser General Public License v2.1
5.01k stars 1.25k forks source link

query a metric, get another metric data. #1237

Open ddqspace-xyz opened 6 years ago

ddqspace-xyz commented 6 years ago

for example:

here is my request: curl http://a.b.c.d:4242/api/query -d '{"start":1529630000000,"end":1529634579000,"queries":[{"aggregator":"sum","metric":"szps_qta_executor.gaia.cluster.app_failed","tags":{"IP": "100.66.33.150"},"downsample":"300s-max-zero"}]}'

here is the response: [{"metric":"szps_skad.k8s.apiserver.apiserver_request_latency","tags":{"IP":"100.66.33.150"},"aggregateTags":["verb","resource"],"dps":{"1529630100":1.4079999853856862,"...}]

as everyone can see, I query data for metric "szps_qta_executor.gaia.cluster.app_failed", but I got another metric "szps_skad.k8s.apiserver.apiserver_request_latency".

biordach commented 6 years ago

Ecnountering the same problem: OpenTSDB version [2.3.0] POST Query body: { "start": "20s-ago", "end": "10s-ago", "queries": [{ "aggregator": "avg", "rate": false, "metric": "trommel.biordach3_raw_touchpoints.value", "tags": {"namespace": "biordach3", "position":"consumer"}, "downsample": "10s-max" }] } Response: [ { "metric": "siphon.open-tsdb-reporter-time-to-build-report.p95", "tags": { "namespace": "biordach3", "position": "Siphon_201", "executorId": "NA", "appHost": "dcos-agentworker-trqe2000003", "appID": "/biordach3/ingest" }, "aggregateTags": [ "partition" ], "dps": { "1531829260": 72235860.33333333 } } ]

Update: restarting the openTSDB instance solved the problem.

manolama commented 5 years ago

This will happen with UID reassignments or deletions or truncations of a table. Make sure to restart a TSD if you modify any UIDs or truncate the UID table so that the UID cache is cleared.

ddqspace-xyz commented 5 years ago

This will happen with UID reassignments or deletions or truncations of a table. Make sure to restart a TSD if you modify any UIDs or truncate the UID table so that the UID cache is cleared.

Is there any other method to fix this issue except resart?