Thanos: query: Inconsistent "thanos_status" metric naming

[MacroPower]

Thanos, Prometheus and Golang version used:

Thanos: 0.11
Promethus: 2.16
Golang: 1.13.6

What happened:

I was exploring some of the internal Thanos metrics, and noticed some inconsistencies with the query component.

Each component has a thanos_status metric, except query, where it is called status.

Results from http://<query>/metrics:

# HELP status Represents status (0 indicates success, 1 indicates failure) of the component.
# TYPE status gauge
status{check="healthy",component="query"} 1
status{check="ready",component="query"} 1

Results from http://<compact/bucket/sidecar/store>/metrics:

# HELP thanos_status Represents status (0 indicates success, 1 indicates failure) of the component.
# TYPE thanos_status gauge
thanos_status{check="healthy",component="sidecar"} 1
thanos_status{check="ready",component="sidecar"} 1

What you expected to happen:

Metric naming should be consistent across all components.

[MacroPower]

0 indicates success, 1 indicates failure

Additionally, this #HELP is the inverse of what it should be.

[jojohappy]

Hi, sorry for the inconsistent metrics, we will fix soon.

0 indicates success, 1 indicates failure

Additionally, this #HELP is the inverse of what it should be.

This HELP message has been changed in this PR #2216 .