A summary metric samples observations over a sliding window of time.
This class represents the metric type summary: https://prometheus.io/docs/instrumenting/writing_clientlibs/#summary
A summary provides a total count of observations and a sum of all observed values. In contrast to a histogram metric it also calculates configurable Phi-quantiles over a sliding window of time.
The essential difference between summaries and histograms is that summaries calculate streaming Phi-quantiles on the client side and expose them directly, while histograms expose bucketed observation counts and the calculation of quantiles from the buckets of a histogram happens on the server side: https://prometheus.io/docs/prometheus/latest/querying/functions/#histogram_quantile.
Note that Phi designates the probability density function of the standard Gaussian distribution.
See https://prometheus.io/docs/practices/histograms/ for detailed explanations of Phi-quantiles, summary usage, and differences to histograms.
The class is thread-safe. No concurrent call to any API of this type causes a data race.
Definition at line 41 of file sources/includes/prometheus/summary.h.
Public Types | |
| using | Quantiles = std::vector< detail::CKMSQuantiles::Quantile > |
| using | Quantiles = std::vector< detail::CKMSQuantiles::Quantile > |
Public Member Functions | |
| Summary (const Quantiles &quantiles, std::chrono::milliseconds max_age=std::chrono::seconds{60}, int age_buckets=5) | |
| Create a summary metric. More... | |
| Summary (Quantiles &&quantiles, std::chrono::milliseconds max_age=std::chrono::seconds{60}, int age_buckets=5) | |
| Create a summary metric. More... | |
| void | Observe (double value) |
| Observe the given amount. More... | |
| ClientMetric | Collect () const |
| Get the current value of the summary. More... | |
| Summary (const Quantiles &quantiles, std::chrono::milliseconds max_age=std::chrono::seconds{60}, int age_buckets=5) | |
| Create a summary metric. More... | |
| Summary (Quantiles &&quantiles, std::chrono::milliseconds max_age=std::chrono::seconds{60}, int age_buckets=5) | |
| Create a summary metric. More... | |
| void | Observe (double value) |
| Observe the given amount. More... | |
| ClientMetric | Collect () const |
| Get the current value of the summary. More... | |
Static Public Attributes | |
| static const MetricType | metric_type {MetricType::Summary} |
| using prometheus::Summary::Quantiles = std::vector<detail::CKMSQuantiles::Quantile> |
Definition at line 43 of file sources/includes/prometheus/summary.h.
| using prometheus::Summary::Quantiles = std::vector<detail::CKMSQuantiles::Quantile> |
Definition at line 43 of file 9.0/sources/includes/prometheus/summary.h.
|
explicit |
Create a summary metric.
| quantiles | A list of 'targeted' Phi-quantiles. A targeted Phi-quantile is specified in the form of a Phi-quantile and tolerated error. For example a Quantile{0.5, 0.1} means that the median (= 50th percentile) should be returned with 10 percent error or a Quantile{0.2, 0.05} means the 20th percentile with 5 percent tolerated error. Note that percentiles and quantiles are the same concept, except percentiles are expressed as percentages. The Phi-quantile must be in the interval [0, 1]. Note that a lower tolerated error for a Phi-quantile results in higher usage of resources (memory and cpu) to calculate the summary. |
The Phi-quantiles are calculated over a sliding window of time. The sliding window of time is configured by max_age and age_buckets.
| max_age | Set the duration of the time window, i.e., how long observations are kept before they are discarded. The default value is 60 seconds. |
| age_buckets | Set the number of buckets of the time window. It determines the number of buckets used to exclude observations that are older than max_age from the summary, e.g., if max_age is 60 seconds and age_buckets is 5, buckets will be switched every 12 seconds. The value is a trade-off between resources (memory and cpu for maintaining the bucket) and how smooth the time window is moved. With only one age bucket it effectively results in a complete reset of the summary each time max_age has passed. The default value is 5. |
|
explicit |
Create a summary metric.
| quantiles | A list of 'targeted' Phi-quantiles. A targeted Phi-quantile is specified in the form of a Phi-quantile and tolerated error. For example a Quantile{0.5, 0.1} means that the median (= 50th percentile) should be returned with 10 percent error or a Quantile{0.2, 0.05} means the 20th percentile with 5 percent tolerated error. Note that percentiles and quantiles are the same concept, except percentiles are expressed as percentages. The Phi-quantile must be in the interval [0, 1]. Note that a lower tolerated error for a Phi-quantile results in higher usage of resources (memory and cpu) to calculate the summary. |
The Phi-quantiles are calculated over a sliding window of time. The sliding window of time is configured by max_age and age_buckets.
| max_age | Set the duration of the time window, i.e., how long observations are kept before they are discarded. The default value is 60 seconds. |
| age_buckets | Set the number of buckets of the time window. It determines the number of buckets used to exclude observations that are older than max_age from the summary, e.g., if max_age is 60 seconds and age_buckets is 5, buckets will be switched every 12 seconds. The value is a trade-off between resources (memory and cpu for maintaining the bucket) and how smooth the time window is moved. With only one age bucket it effectively results in a complete reset of the summary each time max_age has passed. The default value is 5. |
|
explicit |
Create a summary metric.
| quantiles | A list of 'targeted' Phi-quantiles. A targeted Phi-quantile is specified in the form of a Phi-quantile and tolerated error. For example a Quantile{0.5, 0.1} means that the median (= 50th percentile) should be returned with 10 percent error or a Quantile{0.2, 0.05} means the 20th percentile with 5 percent tolerated error. Note that percentiles and quantiles are the same concept, except percentiles are expressed as percentages. The Phi-quantile must be in the interval [0, 1]. Note that a lower tolerated error for a Phi-quantile results in higher usage of resources (memory and cpu) to calculate the summary. |
The Phi-quantiles are calculated over a sliding window of time. The sliding window of time is configured by max_age and age_buckets.
| max_age | Set the duration of the time window, i.e., how long observations are kept before they are discarded. The default value is 60 seconds. |
| age_buckets | Set the number of buckets of the time window. It determines the number of buckets used to exclude observations that are older than max_age from the summary, e.g., if max_age is 60 seconds and age_buckets is 5, buckets will be switched every 12 seconds. The value is a trade-off between resources (memory and cpu for maintaining the bucket) and how smooth the time window is moved. With only one age bucket it effectively results in a complete reset of the summary each time max_age has passed. The default value is 5. |
|
explicit |
Create a summary metric.
| quantiles | A list of 'targeted' Phi-quantiles. A targeted Phi-quantile is specified in the form of a Phi-quantile and tolerated error. For example a Quantile{0.5, 0.1} means that the median (= 50th percentile) should be returned with 10 percent error or a Quantile{0.2, 0.05} means the 20th percentile with 5 percent tolerated error. Note that percentiles and quantiles are the same concept, except percentiles are expressed as percentages. The Phi-quantile must be in the interval [0, 1]. Note that a lower tolerated error for a Phi-quantile results in higher usage of resources (memory and cpu) to calculate the summary. |
The Phi-quantiles are calculated over a sliding window of time. The sliding window of time is configured by max_age and age_buckets.
| max_age | Set the duration of the time window, i.e., how long observations are kept before they are discarded. The default value is 60 seconds. |
| age_buckets | Set the number of buckets of the time window. It determines the number of buckets used to exclude observations that are older than max_age from the summary, e.g., if max_age is 60 seconds and age_buckets is 5, buckets will be switched every 12 seconds. The value is a trade-off between resources (memory and cpu for maintaining the bucket) and how smooth the time window is moved. With only one age bucket it effectively results in a complete reset of the summary each time max_age has passed. The default value is 5. |
| ClientMetric prometheus::Summary::Collect | ( | ) | const |
Get the current value of the summary.
Collect is called by the Registry when collecting metrics.
| ClientMetric prometheus::Summary::Collect | ( | ) | const |
Get the current value of the summary.
Collect is called by the Registry when collecting metrics.
| void prometheus::Summary::Observe | ( | double | value | ) |
Observe the given amount.
| void prometheus::Summary::Observe | ( | double | value | ) |
Observe the given amount.
|
static |
Definition at line 45 of file sources/includes/prometheus/summary.h.