The operator changes the meaning of the metric, so the metric name should
be dropped. Technically this would be a breaking change, but it's also very
obviously a bug and not likely that anyone depends on it.
Signed-off-by: Julius Volz <julius.volz@gmail.com>
The bounds weren't really used so far, so no actual bug in the code so
far. But it's obviously confusing if the bounds returned by a
floatBucketIterator with a target schema different from the original
schema are wrong.
Signed-off-by: beorn7 <beorn@grafana.com>
If a float histogram has a zero bucket with a threshold of zero _and_
an empty zero bucket, it wasn't identified as a native histogram
because the `isNativeHistogram` helper function only looked at integer
buckets.
Signed-off-by: beorn7 <beorn@grafana.com>
Native histograms without a zero threshold aren't federated properly.
This adds a test to prove the specific failure mode, which is that
histograms with a zero threshold of zero are federated as classic
histograms.
The underlying reason is that the protobuf parser identifies a native
histogram by detecting a zero bucket or by detecting integer buckets.
Therefore, a float histogram with a zero threshold of zero and an
unpopulated zero bucket falls through the cracks (no integer buckets,
no zero bucket).
This commit also addse a test case for the latter.
Signed-off-by: beorn7 <beorn@grafana.com>
Version 2 introduced a breaking change in the `id` field of all
resources. They were changed from `int` to `int64` to make sure that all
future numerical IDs are supported on all architectures.
You can learn more about this
[here](https://docs.hetzner.cloud/#deprecation-notices-%E2%9A%A0%EF%B8%8F)
Signed-off-by: Julian Tölle <julian.toelle@hetzner-cloud.de>
InstanceSpec struct members are untyped integers, so they can overflow
on 32-bit arch when bit-shifted left.
Signed-off-by: Daniel Swarbrick <daniel.swarbrick@gmail.com>
Continue to remove confusion that histogram samples are also samples
and histogram values are also values etc. by renaming float values and
float samples using the same schema as for histograms.
Concretely:
- result → resultFloats (corresponding to resultHistograms)
- pendingResult → pendingFloats (corresponding to pendingHistograms)
- rolledbackResult → rolledbackFloats (corresponding to rolledbackHistograms)
- sample → floatSample (corresponding to histogramSample)
This also order the fields in `collectResultAppender` more
consistently.
Signed-off-by: beorn7 <beorn@grafana.com>
This has become a requirement for native histograms, as a single
histogram sample commonly has many buckets, so that providing many
exemplars makes sense.
Since OM text doesn't support native histograms yet, the test had to
be expanded to also support protobuf test cases.
Signed-off-by: beorn7 <beorn@grafana.com>
The problem was the following:
When trying to parse native histograms and classic histograms in
parallel, the parser would first parse the histogram proto messages as
a native histogram and then parse the same message again, but now as a
classic histogram. Afterwards, it would forget that it was dealing
with a metric family that contains native histograms and would parse
the rest of the metric family as classic histograms only. The fix is
to check again after being done with a classic histogram.
Signed-off-by: beorn7 <beorn@grafana.com>
* Add function for iterating through all buckets in reverse to find max bucket
Signed-off-by: Carrie Edwards <edwrdscarrie@gmail.com>
* enhance histogram_quantile to get min/max value
Signed-off-by: Ziqi Zhao <zhaoziqi9146@gmail.com>
---------
Signed-off-by: Carrie Edwards <edwrdscarrie@gmail.com>
Signed-off-by: Ziqi Zhao <zhaoziqi9146@gmail.com>
Co-authored-by: Carrie Edwards <edwrdscarrie@gmail.com>
Snappy remains as the default compression but there is now a flag to switch
the compression algorithm.
Signed-off-by: Justin Lei <justin.lei@grafana.com>
