PR #12557 introduced the possibility of parsing multiple exemplars per
native histograms. It did so by requiring the `Exemplar` method of the
parser to be called repeatedly until it returns false. However, the
protobuf parser code wasn't correctly updated for the old case of a
single exemplar for a classic bucket (if actually parsed as a classic
bucket) and a single exemplar on a counter. In those cases, the method
would return `true` forever, yielding the same exemplar again and
again, leading to an endless loop.
With this fix, the state is now tracked and the single exemplar is
only returned once.
Signed-off-by: beorn7 <beorn@grafana.com>
Native histograms without observations and with a zero threshold of
zero look the same as classic histograms in the protobuf exposition
format. According to
https://github.com/prometheus/client_golang/issues/1127 , the idea is
to add a no-op span to those histograms to mark them as native
histograms. This commit enables Prometheus to detect that no-op span
and adds a doc comment to the proto spec describing the behavior.
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>
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>
So far, if a target exposes a histogram with both classic and native
buckets, a native-histogram enabled Prometheus would ignore the
classic buckets. With the new scrape config option
`scrape_classic_histograms` set, both buckets will be ingested,
creating all the series of a classic histogram in parallel to the
native histogram series. For example, a histogram `foo` would create a
native histogram series `foo` and classic series called `foo_sum`,
`foo_count`, and `foo_bucket`.
This feature can be used in a migration strategy from classic to
native histograms, where it is desired to have a transition period
during which both native and classic histograms are present.
Note that two bugs in classic histogram parsing were found and fixed
as a byproduct of testing the new feature:
1. Series created from classic _gauge_ histograms didn't get the
_sum/_count/_bucket prefix set.
2. Values of classic _float_ histograms weren't parsed properly.
Signed-off-by: beorn7 <beorn@grafana.com>
If a (float or integer) histogram is a gauge histogram, set the
CounterResetHint accordingly. (The default value is fine for the
normal counter histograms.)
Signed-off-by: beorn7 <beorn@grafana.com>
With this commit, the parser stops to see a gauge histogram (whether
native or conventional) as an unexpected metric type. It ingests it
normally, it even sets the `GaugeHistogram` type in the metadata (as
it has already done for a conventional gauge histogram scraped using
OpenMetrics), but it otherwise treats it as a normal counter-like
histogram.
Once #11783 is merged, though, it should be very easy to utilize the
type information.
Signed-off-by: beorn7 <beorn@grafana.com>
So far, the parser hasn't validated that the type is valid in the
`Next()` call. Later, in the `Series()` call, however, it assumes that
we will only see valid types and therefore panics with `encountered
unexpected metric type, this is a bug`.
This commit fixes said bug by adding validation to the `Next()` call.
Signed-off-by: beorn7 <beorn@grafana.com>
And use the new method to call to compact Histograms during
parsing. This happens for both `Histogram` and `FloatHistogram`. In
this way, if targets decide to optimize the exposition size by merging
spans with empty buckets in between, we still get a normalized
results. It will also normalize away any valid but weird
representations like empty spans, spans with offset zero, and empty
buckets at the start or end of a span.
The implementation seemed easy at first as it just turns the
`compactBuckets` helper into a generic function (which now got its own
file). However, the integer Histograms have delta buckets instead of
absolute buckets, which had to be treated specially in the generic
`compactBuckets` function. To make sure it works, I have added plenty
of explicit tests for `Histogram` in addition to the `FloatHistogram`
tests.
I have also updated the doc comment for the `Compact` method.
Based on the insights now expressed in the doc comment, compacting
with a maxEmptyBuckets > 0 is rarely useful. Therefore, this commit
also sets the value to 0 in the two cases we were using 3 so far. We
might still want to reconsider, so I don't want to remove the
maxEmptyBuckets parameter right now.
Signed-off-by: beorn7 <beorn@grafana.com>
