Documentation update with new metrics

README.md

@@ -519,17 +519,26 @@ Where:
 * `REASON` is an arbitrary error reason term (in case of `status="error"`) or an empty string (when `status="success"`)
 * `LE` defines the `upper inclusive bound` (`less than or equal`) values for buckets, currently `1000`, `10_000`, `25_000`, `50_000`, `100_000`, `250_000`, `500_000`, `1000_000` or `+Inf`
 
+This histogram metric shows the distribution of times needed to:
+1. Select a worker (this may include waiting time when all workers are busy).
+2. Send a request.
+3. Get a response from push notifications provider.
+
+###### HTTP/2 requests
+
+`sparrow_h_worker_handle_duration_microsecond_bucket{le=${LE}}`
+`sparrow_h_worker_handle_duration_microsecond_sum{le=${LE}}`
+`sparrow_h_worker_handle_duration_microsecond_count{le=${LE}}`
+
+Where:
+* `LE` defines the `upper inclusive bound` (`less than or equal`) values for buckets, currently `1000`, `10_000`, `25_000`, `50_000`, `100_000`, `250_000`, `500_000`, `1000_000` or `+Inf`
+
 > **NOTE**
 >
 > A bucket of value 250_000 will keep the count of measurements that are less than or equal to 250_000.
 > A measurement of value 51_836 will be added to all the buckets where the upper bound is greater than 51_836.
 > In this case these are buckets `100_000`, `250_000`, `500_000`, `1000_000` and `+Inf`
 
-This histogram metric shows the distribution of times needed to:
-1. Select a worker (this may include waiting time when all workers are busy).
-2. Send a request.
-3. Get a response from push notifications provider.
-
 ##### Counters
 
 * `mongoose_push_supervisor_init_count{service=${SERVICE}}` - Counts the number of push notification service supervisor starts.
@@ -538,6 +547,8 @@ This histogram metric shows the distribution of times needed to:
 * `mongoose_push_apns_state_init_count` - Counts the number of APNS state initialisations.
 * `mongoose_push_apns_state_terminate_count` - Counts the number of APNS state terminations.
 * `mongoose_push_apns_state_get_default_topic_count` - Counts the number of default topic reads from cache.
+* `sparrow_pools_warden_pools_count` - Counts the number of worker pools.
+* `sparrow_pools_warden_workers_count{pool=${POOL}}` - Counts the number of workers operated by a given worker `POOL`.
 
 #### How to quickly see all metrics
 

lib/mongoose_push/metrics/telemetry_metrics.ex

@@ -48,7 +48,9 @@ defmodule MongoosePush.Metrics.TelemetryMetrics do
         "sparrow.h2_worker.handle.duration.microsecond",
         event_name: [:sparrow, :h2_worker, :handle],
         measurement: :time,
-        buckets: [10_000, 25_000, 50_000, 100_000, 200_000, 500_000, 1000_000],
+        reporter_options: [
+          buckets: [10_000, 25_000, 50_000, 100_000, 200_000, 500_000, 1000_000]
+        ],
         description: "A histogram showing time it takes for h2_worker to handle request."
       ),
       Telemetry.Metrics.last_value(

test/unit/mongoose_push_telemetry_metrics_test.exs

@@ -130,7 +130,7 @@ defmodule MongoosePushTelemetryMetricsTest do
       ~r/mongoose_push_notification_send_time_microsecond_count{error_category=\"\",error_reason=\"\",service=\"fcm\",status=\"success\"} (?<count>[\d]+)/
 
     fcm_match = Regex.named_captures(fcm_regex, metrics.resp_body)
-    assert 0 != fcm_match
+    assert nil != fcm_match
   end
 
   test "sparrow periodic metrics" do