VictoriaMetrics/docs/stream-aggregation.md

---
sort: 98
---

# streaming aggregation

[vmagent](https://docs.victoriametrics.com/vmagent.html) and [single-node VictoriaMetrics](https://docs.victoriametrics.com/Single-server-VictoriaMetrics.html)
can aggregate incoming [samples](https://docs.victoriametrics.com/keyConcepts.html#raw-samples) in streaming mode by time and by labels.
The aggregation is applied to all the metrics received via any [supported data ingestion protocol](https://docs.victoriametrics.com/#how-to-import-time-series-data)
and/or scraped from [Prometheus-compatible targets](https://docs.victoriametrics.com/#how-to-scrape-prometheus-exporters-such-as-node-exporter).

The stream aggregation is configured via the following command-line flags:

- `-remoteWrite.streamAggr.config` at [vmagent](https://docs.victoriametrics.com/vmagent.html).
  This flag can be specified individually per each specified `-remoteWrite.url`.
  This allows writing different aggregates to different remote storage destinations.
- `-streamAggr.config` at [single-node VictoriaMetrics](https://docs.victoriametrics.com/Single-server-VictoriaMetrics.html).

These flags must point to a file containing [stream aggregation config](#stream-aggregation-config).

By default only the aggregated data is written to the storage. If the original incoming samples must be written to the storage too,
then the following command-line flags must be specified:

- `-remoteWrite.streamAggr.keepInput` at [vmagent](https://docs.victoriametrics.com/vmagent.html).
  This flag can be specified individually per each specified `-remoteWrite.url`.
  This allows writing both raw and aggregate data to different remote storage destinations.
- `-streamAggr.keepInput` at [single-node VictoriaMetrics](https://docs.victoriametrics.com/Single-server-VictoriaMetrics.html).

Stream aggregation ignores timestamps associated with the input [samples](https://docs.victoriametrics.com/keyConcepts.html#raw-samples).
It expects that the ingested samples have timestamps close to the current time.

## Use cases

Stream aggregation can be used in the following cases:

* [Statsd alternative](#statsd-alternative)
* [Recording rules alternative](#recording-rules-alternative)
* [Reducing the number of stored samples](#reducing-the-number-of-stored-samples)
* [Reducing the number of stored series](#reducing-the-number-of-stored-series)

### Statsd alternative

Stream aggregation can be used as [statsd](https://github.com/statsd/statsd) altnernative in the following cases:

* [Counting input samples](#counting-input-samples)
* [Summing input metrics](#summing-input-metrics)
* [Quantiles over input metrics](#quantiles-over-input-metrics)
* [Histograms over input metrics](#histograms-over-input-metrics)

### Recording rules alternative

Sometimes [alerting queries](https://docs.victoriametrics.com/vmalert.html#alerting-rules) may require non-trivial amounts of CPU, RAM,
disk IO and network bandwith at metrics storage side. For example, if `http_request_duration_seconds` histogram is generated by thousands
of app instances, then the alerting query `histogram_quantile(0.99, sum(increase(http_request_duration_seconds_bucket[5m])) without (instance)) > 0.5`
can become slow, since it needs to scan too big number of unique [time series](https://docs.victoriametrics.com/keyConcepts.html#time-series)
with `http_request_duration_seconds_bucket` name. This alerting query can be sped up by pre-calculating
the `sum(increase(http_request_duration_seconds_bucket[5m])) without (instance)` via [recording rule](https://docs.victoriametrics.com/vmalert.html#recording-rules).
But this recording rule may take too much time to execute too. In this case the slow recording rule can be substituted
with the following [stream aggregation config](#stream-aggregation-config):

```yaml
- match: 'http_request_duration_seconds_bucket'
  interval: 5m
  without: [instance]
  outputs: [total]
```

This stream aggregation generates `http_request_duration_seconds_bucket:5m_without_instance_total` output series according to [output metric naming](#output-metric-names).
Then these series can be used in [alerting rules](https://docs.victoriametrics.com/vmalert.html#alerting-rules):

```metricsql
histogram_quantile(0.99, last_over_time(http_request_duration_seconds_bucket:5m_without_instance_total[5m])) > 0.5
```

This query is executed much faster than the original query, because it needs to scan much lower number of time series.

See [the list of aggregate output](#aggregation-outputs), which can be specified at `output` field.
See also [aggregating by labels](#aggregating-by-labels).


### Reducing the number of stored samples

If per-[series](https://docs.victoriametrics.com/keyConcepts.html#time-series) samples are ingested at high frequency,
then this may result in high disk space usage, since too much data must be stored to disk. This also may result
in slow queries, since too much data must be processed during queries.

This can be fixed with the stream aggregation by increasing the interval between per-series samples stored in the database.

For example, the following [stream aggregation config](#stream-aggregation-config) reduces the frequency of input samples
to one sample per 5 minutes per each input time series (this operation is also known as downsampling):

```yaml
  # Aggregate metrics ending with _total with `total` output.
  # See https://docs.victoriametrics.com/stream-aggregation.html#aggregation-outputs
- match: '{__name__=~".+_total"}'
  interval: 5m
  outputs: [total]

  # Downsample other metrics with `count_samples`, `sum_samples`, `min` and `max` outputs
  # See https://docs.victoriametrics.com/stream-aggregation.html#aggregation-outputs
- match: '{__name__!~".+_total"}'
  interval: 5m
  outputs: [count_samples, sum_samples, min, max]
```

The aggregated output metrics have the following names according to [output metric naming](#output-metric-names):

```
# For input metrics ending with _total
some_metric_total:5m_total

# For input metrics not ending with _total
some_metric:5m_count_samples
some_metric:5m_sum_samples
some_metric:5m_min
some_metric:5m_max
```

See [the list of aggregate output](#aggregation-outputs), which can be specified at `output` field.
See also [aggregating by labels](#aggregating-by-labels).

### Reducing the number of stored series

Sometimes apps may generate too many [time series](https://docs.victoriametrics.com/keyConcepts.html#time-series).
For example, the `http_requests_total` metric may have `path` or `user` label with too big number of unique values.
In this case the following stream aggregation can be used for reducing the number metrics stored in VictoriaMetrics:

```yaml
- match: 'http_requests_total'
  interval: 30s
  without: [path, user]
  outputs: [total]
```

This config specifies labels, which must be removed from the aggregate outpit, in the `without` list.
See [these docs](#aggregating-by-labels) for more details.

The aggregated output metric has the following name according to [output metric naming](#output-metric-names):

```
http_requests_total:30s_without_path_user_total
```

See [the list of aggregate output](#aggregation-outputs), which can be specified at `output` field.


### Counting input samples

If the monitored app generates event-based metrics, then it may be useful to count the number of such metrics
at stream aggregation level.

For example, if an advertising server generates `hits{some="labels"} 1` and `clicks{some="labels"} 1` metrics
per each incoming hit and click, then the following [stream aggregation config](#stream-aggregation-config)
can be used for counting these metrics per every 30 second interval:

```yml
- match: '{__name__=~"hits|clicks"}'
  interval: 30s
  outputs: [count_samples]
```

This config generates the following output metrics for `hits` and `clicks` input metrics
according to [output metric naming](#output-metric-names):

```
hits:30s_count_samples count1
clicks:30s_count_samples count2
```

See [the list of aggregate output](#aggregation-outputs), which can be specified at `output` field.
See also [aggregating by labels](#aggregating-by-labels).


### Summing input metrics

If the monitored app calulates some events and then sends the calculated number of events to VictoriaMetrics
at irregular intervals or at too high frequency, then stream aggregation can be used for summing such events
and writing the aggregate sums to the storage at regular intervals.

For example, if an advertising server generates `hits{some="labels} N` and `clicks{some="labels"} M` metrics
at irregular intervals, then the following [stream aggregation config](#stream-aggregation-config)
can be used for summing these metrics per every minute:

```yml
- match: '{__name__=~"hits|clicks"}'
  interval: 1m
  outputs: [sum_samples]
```

This config generates the following output metrics according to [output metric naming](#output-metric-names):

```
hits:1m_sum_samples sum1
clicks:1m_sum_samples sum2
```

See [the list of aggregate output](#aggregation-outputs), which can be specified at `output` field.
See also [aggregating by labels](#aggregating-by-labels).


### Quantiles over input metrics

If the monitored app generates measurement metrics per each request, then it may be useful to calculate
the pre-defined set of [percentiles](https://en.wikipedia.org/wiki/Percentile) over these measurements.

For example, if the monitored app generates `request_duration_seconds N` and `response_size_bytes M` metrics
per each incoming request, then the following [stream aggregation config](#stream-aggregation-config)
can be used for calculating 50th and 99th percentiles for these metrics every 30 seconds:

```yaml
- match: '{__name__=~"request_duration_seconds|response_size_bytes"}'
  interval: 30s
  outputs: ["quantiles(0.50, 0.99)"]
```

This config generates the following output metrics according to [output metric naming](#output-metric-names):

```
request_duration_seconds:30s_quantiles{quantile="0.50"} value1
request_duration_seconds:30s_quantiles{quantile="0.99"} value2

response_size_bytes:30s_quantiles{quantile="0.50"} value1
response_size_bytes:30s_quantiles{quantile="0.99"} value2
```

See [the list of aggregate output](#aggregation-outputs), which can be specified at `output` field.
See also [histograms over input metrics](#histograms-over-input-metrics) and [aggregating by labels](#aggregating-by-labels).

### Histograms over input metrics

If the monitored app generates measurement metrics per each request, then it may be useful to calculate
a [histogram](https://docs.victoriametrics.com/keyConcepts.html#histogram) over these metrics.

For example, if the monitored app generates `request_duration_seconds N` and `response_size_bytes M` metrics
per each incoming request, then the following [stream aggregation config](#stream-aggregation-config)
can be used for calculating [VictoriaMetrics histogram buckets](https://valyala.medium.com/improving-histogram-usability-for-prometheus-and-grafana-bc7e5df0e350)
for these metrics every 60 seconds:

```yaml
- match: '{__name__=~"request_duration_seconds|response_size_bytes"}'
  interval: 60s
  outputs: [histogram_bucket]
```

This config generates the following output metrics according to [output metric naming](#output-metric-names).

```
request_duration_seconds:60s_histogram_bucket{vmrange="start1...end1"} count1
request_duration_seconds:60s_histogram_bucket{vmrange="start2...end2"} count2
...
request_duration_seconds:60s_histogram_bucket{vmrange="startN...endN"} countN

response_size_bytes:60s_histogram_bucket{vmrange="start1...end1"} count1
response_size_bytes:60s_histogram_bucket{vmrange="start2...end2"} count2
...
response_size_bytes:60s_histogram_bucket{vmrange="startN...endN"} countN
```

The resulting histogram buckets can be queried with [MetricsQL](https://docs.victoriametrics.com/MetricsQL.html) in the following ways:

1. An estimated 50th and 99th [percentiles](https://en.wikipedia.org/wiki/Percentile) of the request duration over the last hour:

   ```metricsql
   histogram_quantiles("quantile", 0.50, 0.99, sum(increase(request_duration_seconds:60s_histogram_bucket[1h])) by (vmrange))
   ```

   This query uses [histogram_quantiles](https://docs.victoriametrics.com/MetricsQL.html#histogram_quantiles) function.

2. An estimated [standard deviation](https://en.wikipedia.org/wiki/Standard_deviation) of the request duration over the last hour:

   ```metricsql
   histogram_stddev(sum(increase(request_duration_seconds:60s_histogram_bucket[1h])) by (vmrange))
   ```

   This query uses [histogram_stddev](https://docs.victoriametrics.com/MetricsQL.html#histogram_stddev) function.

3. An estimated share of requests with the duration smaller than `0.5s` over the last hour:

   ```metricsql
   histogram_share(0.5, sum(increase(request_duration_seconds:60s_histogram_bucket[1h])) by (vmrange))
   ```

   This query uses [histogram_share](https://docs.victoriametrics.com/MetricsQL.html#histogram_share) function.

See [the list of aggregate output](#aggregation-outputs), which can be specified at `output` field.
See also [quantiles over input metrics](#quantiles-over-input-metrics) and [aggregating by labels](#aggregating-by-labels).


## Output metric names

Output metric names for stream aggregation are constructed according to the following pattern:

```
<metric_name>:<interval>[_by_<by_labels>][_without_<without_labels>]_<output>
```

- `<metric_name>` is the original metric name.
- `<interval>` is the interval specified in the [stream aggregation config](#stream-aggregation-config).
- `<by_labels>` is `_`-delimited list of `by` labels specified in the [stream aggregation config](#stream-aggregation-config).
  If the `by` list is missing in the config, then the `_by_<by_labels>` part isn't included in the output metric name.
- `<without_labels>` is an optional `_`-delimited list of `without` labels specified in the [stream aggregation config](#stream-aggregation-config).
  If the `without` list is missing in the config, then the `_without_<without_labels>` part isn't included in the output metric name.
- `<output>` is the aggregate used for constucting the output metric. The aggregate name is taken from the `outputs` list
  at the corresponding [stream aggregation config](#stream-aggregation-config).

Both input and ouput metric names can be modified if needed via relabeling according to [these docs](#relabeling).


## Relabeling

It is possible to apply [arbitrary relabeling](https://docs.victoriametrics.com/vmagent.html#relabeling) to input and output metrics
during stream aggregation via `input_relabel_configs` and `output_relabel_config` options in [stream aggregation config](#stream-aggregation-config).

For example, the following config removes the `:1m_sum_samples` suffix added [to the output metric name](#output-metric-names):

```yml
- interval: 1m
  outputs: [sum_samples]
  output_relabel_configs:
  - source_labels: [__name__]
    target_label: __name__
    regex: "(.+):.+"
```

## Aggregation outputs

The following aggregation outputs are supported in the `outputs` list of the [stream aggregation config](#stream-aggregation-config):

* `total` generates output [counter](https://docs.victoriametrics.com/keyConcepts.html#counter) by summing the input counters.
  The `total` handler properly handles input counter resets.
  The `total` handler returns garbage when something other than [counter](https://docs.victoriametrics.com/keyConcepts.html#counter) is passed to the input.
* `increase` returns the increase of input [counters](https://docs.victoriametrics.com/keyConcepts.html#counter).
  The `increase` handler properly handles the input counter resets.
  The `increase` handler returns garbage when something other than [counter](https://docs.victoriametrics.com/keyConcepts.html#counter) is passed to the input.
* `count_series` counts the number of unique [time series](https://docs.victoriametrics.com/keyConcepts.html#time-series).
* `count_samples` counts the number of input [samples](https://docs.victoriametrics.com/keyConcepts.html#raw-samples).
* `sum_samples` sums input [sample values](https://docs.victoriametrics.com/keyConcepts.html#raw-samples).
* `last` returns the last input [sample value](https://docs.victoriametrics.com/keyConcepts.html#raw-samples).
* `min` returns the minimum input [sample value](https://docs.victoriametrics.com/keyConcepts.html#raw-samples).
* `max` returns the maximum input [sample value](https://docs.victoriametrics.com/keyConcepts.html#raw-samples).
* `avg` returns the average input [sample value](https://docs.victoriametrics.com/keyConcepts.html#raw-samples).
* `stddev` returns [standard deviation](https://en.wikipedia.org/wiki/Standard_deviation) for the input [sample values](https://docs.victoriametrics.com/keyConcepts.html#raw-samples).
* `stdvar` returns [standard variance](https://en.wikipedia.org/wiki/Variance) for the input [sample values](https://docs.victoriametrics.com/keyConcepts.html#raw-samples).
* `histogram_bucket` returns [VictoriaMetrics histogram buckets](https://valyala.medium.com/improving-histogram-usability-for-prometheus-and-grafana-bc7e5df0e350)
  for the input [sample values](https://docs.victoriametrics.com/keyConcepts.html#raw-samples).
* `quantiles(phi1, ..., phiN)` returns [percentiles](https://en.wikipedia.org/wiki/Percentile) for the given `phi*`
  over the input [sample values](https://docs.victoriametrics.com/keyConcepts.html#raw-samples).
  The `phi` must be in the range `[0..1]`, where `0` means `0th` percentile, while `1` means `100th` percentile.

The aggregations are calculated during the `interval` specified in the [config](#stream-aggregation-config)
and then sent to the storage.

If `by` and `without` lists are specified in the [config](#stream-aggregation-config),
then the [aggregation by labels](#aggregating-by-labels) is performed additionally to aggregation by `interval`.


## Aggregating by labels

All the labels for the input metrics are preserved by default in the output metrics. For example,
the input metric `foo{app="bar",instance="host1"}` results to the output metric `foo:1m_sum_samples{app="bar",instance="host1"}`
when the following [stream aggregation config](#stream-aggregation-config) is used:

```yaml
- interval: 1m
  outputs: [sum_samples]
```

The input labels can be removed via `without` list specified in the config. For example, the following config
removes the `instance` label from output metrics by summing input samples across all the instances:

```yaml
- interval: 1m
  without: [instance]
  outputs: [sum_samples]
```

In this case the `foo{app="bar",instance="..."}` input metrics are transformed into `foo:1m_without_instance_sum_samples{app="bar"}`
output metric.

It is possible specifying the exact list of labels in the output metrics via `by` list.
For example, the following config sums input samples by the `app` label:

```yaml
- interval: 1m
  by: [app]
  outputs: [sum_samples]
```

In this case the `foo{app="bar",instance="..."}` input metrics are transformed into `foo:1m_by_app_sum_samples{app="bar"}`
output metric.


## Stream aggregation config

Below is the format for stream aggregation config file, which may be referred via `-remoteWrite.streamAggr.config` command-line flag
at [vmagent](https://docs.victoriametrics.com/vmagent.html) or via `-streamAggr.config` command-line flag
at [single-node VictoriaMetrics](https://docs.victoriametrics.com/Single-server-VictoriaMetrics.html):

```yaml
  # match is an optional filter for incoming samples to aggregate.
  # It can contain arbitrary Prometheus series selector
  # according to https://docs.victoriametrics.com/keyConcepts.html#filtering .
  # If match is missing, then all the incoming samples are aggregated.
- match: 'http_request_duration_seconds_bucket{env=~"prod|staging"}'

  # interval is the interval for the aggregation.
  # The aggregated stats is sent to remote storage once per interval.
  interval: 1m

  # without is an optional list of labels, which must be removed from the output aggregation.
  # See https://docs.victoriametrics.com/stream-aggregation.html#aggregating-by-labels
  without: [instance]

  # by is an optioanl list of labels, which must be preserved in the output aggregation.
  # See https://docs.victoriametrics.com/stream-aggregation.html#aggregating-by-labels
  # by: [job, vmrange]

  # outputs is the list of aggregations to perform on the input data.
  # See https://docs.victoriametrics.com/stream-aggregation.html#aggregation-outputs
  outputs: [total]

  # input_relabel_configs is an optional relabeling rules,
  # which are applied to the incoming samples after they pass the match filter
  # and before being aggregated.
  # See https://docs.victoriametrics.com/stream-aggregation.html#relabeling
  input_relabel_configs:
  - target_label: vmaggr
    replacement: before

  # output_relabel_configs is an optional relabeling rules,
  # which are applied to the aggregated output metrics.
  output_relabel_configs:
  - target_label: vmaggr
    replacement: after
```

The file can contain multiple aggregation configs. The aggregation is performed independently
per each specified config entry.