feat: Allow users to register the same Meter multiple times without exception

[mayurkale22]

No description provided.

[codecov-io]

Codecov Report

Merging #2017 into master will increase coverage by <.01%.
The diff coverage is 55%.

@@             Coverage Diff              @@
##             master    #2017      +/-   ##
============================================
+ Coverage     83.11%   83.11%   +<.01%     
- Complexity     2249     2252       +3     
============================================
  Files           319      319              
  Lines         10527    10529       +2     
  Branches       1047     1048       +1     
============================================
+ Hits           8749     8751       +2     
- Misses         1438     1440       +2     
+ Partials        340      338       -2

Impacted Files	Coverage Δ	Complexity Δ
...ensus/implcore/metrics/DerivedDoubleGaugeImpl.java	94.91% <0%> (-1.64%)	11 <0> (ø)
...us/implcore/metrics/DerivedLongCumulativeImpl.java	95.45% <0%> (-1.47%)	11 <0> (ø)
...ncensus/implcore/metrics/DoubleCumulativeImpl.java	93.33% <0%> (-1.27%)	16 <0> (ø)
...ncensus/implcore/metrics/DerivedLongGaugeImpl.java	91.52% <0%> (-1.58%)	10 <0> (ø)
...pencensus/implcore/metrics/LongCumulativeImpl.java	93.33% <0%> (-1.27%)	16 <0> (ø)
.../implcore/metrics/DerivedDoubleCumulativeImpl.java	95.45% <0%> (-1.47%)	11 <0> (ø)
...o/opencensus/implcore/metrics/DoubleGaugeImpl.java	94.44% <100%> (+0.07%)	17 <1> (+1)	⬆️
.../io/opencensus/implcore/metrics/LongGaugeImpl.java	94.44% <100%> (+0.07%)	17 <1> (+1)	⬆️
...pencensus/implcore/metrics/MetricRegistryImpl.java	72.82% <75%> (+1.39%)	7 <0> (ø)	⬇️
...re/trace/export/InProcessSampledSpanStoreImpl.java	96.44% <0%> (+0.59%)	22% <0%> (ø)	⬇️
... and 1 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update b6d4cbd...cc9073d. Read the comment docs.

[rghetia]

LGTM except for one minor comment.

[rghetia]

As discussed, please add test for

m1 = register.AddGauge("g1")
m2 = register.AddGauge("g1")

expect m1 and m2 to be the same.

[mayurkale22]

@rghetia please review and approve when you get a chance, thank you :)

[olavloite]

(Note, this comment is not relevant for this PR, and is only intended for discussing whether this fix would also fix this specific issue for the Spanner client library).

@mayurkale22 Can you confirm that registering multiple metrics with different label values (but equal label keys) will not go down this code path?

[mayurkale22]

Can you confirm that registering multiple metrics with different label values (but equal label keys) will not go down this code path?

Yes, it is normal to have different label values - those will be represented as individual TimeSeries in the metric. i.e:

Metric:
     {name, label_keys, description, unit, type, ...}
TimeSeries
    [{labels_values1, value1}, {labels_values2, value2}, ...]

See: https://github.com/census-instrumentation/opencensus-java/blob/master/api/src/main/java/io/opencensus/metrics/export/Metric.java#L51

The exception will be raised only when you use the same metric name but with a different type or unit or label_keys.

[olavloite]

That sounds good 👍 Thanks for looking into this.

I think this fix is better than what I added to the Spanner client library, so I'll close the PR.

The only concern that I still have is what it looks like in the dashboards when there are multiple different instances of a SessionPool with the same label values. I understand that it won't cause any errors in the client library, but how is it presented to the user? This situation would occur when an application is opening two connections to the same database with different SessionPoolOptions. This would also mean different values for constant metrics like MaxSessions.