Histogram Aggregator break metric expiration inside Prometheus client output

[opengamer29]

Relevent telegraf.conf

[agent]
  interval = "30s"
  round_interval = true
  metric_batch_size = 20000
  metric_buffer_limit = 10000
  collection_jitter = "0s"
  flush_jitter = "0s"
  debug = true
  quiet = false

[[inputs.socket_listener]]
  service_address = "udp://:8125"
  data_format = "influx"

[[outputs.prometheus_client]]
  listen = ":9201"
  expiration_interval = "30s"
  metric_version = 2

[[aggregators.histogram]]
  period = "30s"
  drop_original = true
  reset = false
  cumulative = true

  [[aggregators.histogram.config]]
    measurement_name = "test"
    buckets = [0.0, 100.0, 300.0, 500.0, 700.0, 1000.0, 1500.0]

System info

Telegraf 1.20.0

Docker

No response

Steps to reproduce

1. Send test ms=1 to udp://:8125
2. Check /metrics endpoint -> buckets for test metric should be shown
3. Wait for more that expiration_interval (30 seconds)
4. Check /metrics endpoint

Expected behavior

There are no metrics inside /metrics endpoint due to its expiration.

Actual behavior

Expired metrics are still presented.

Additional info

It's important for cases described in #10517 It allows to significantly decrease service loading.
Following the code, Histogram Aggregator always pushes metrics with the current timestamp (historgram.go:173).