PromQL bug with query_range and metrics name change #4562
Description
Bug Report
What did you do?
- Upgrade node_exporter
- run with query_range:
sum(rate({__name__=~'node_network_receive_drop_total|node_network_receive_drop',env='prod',instance="nip-beprd51"}[5m]) > 0) by(instance,cluster)
What did you expect to see?
a constant result
What did you see instead? Under which circumstances?
when the query range overlaps the metric name change, I git no result when we switch to the new metric
Environment
- Prometheus version:
prometheus, version 2.3.2 (branch: HEAD, revision: 71af5e2)
build user: root@5258e0bd9cc1
build date: 20180712-14:02:52
go version: go1.10.3
Raw data
{"status":"success","data":{"resultType":"matrix","result":[{"metric":{"__name__":"node_network_receive_drop","cluster":"default","device":"bond0","env":"prod","instance":"nip-beprd51","job":"node_exporters"},"values":[[1535541130.395,"107315172"],[1535541160.395,"107315217"]]},{"metric":{"__name__":"node_network_receive_drop_total","cluster":"default","device":"bond0","env":"prod","instance":"nip-beprd51","job":"node_exporters"},"values":[[1535541190.396,"107315254"],[1535541220.395,"107315281"],[1535541250.395,"107315334"],[1535541280.395,"107315405"],[1535541310.395,"107315457"],[1535541340.395,"107315529"],[1535541370.395,"107315598"],[1535541400.395,"107315649"],[1535541430.395,"107315697"]]}]}}
Query range queries with sum() that cross 1535541430-1535541431 are stopped in the middle
