Metrics Endpoint
The Metrics endpoint is used to access runtime properties of the running node. It exposes functionality provided by the Dropwizard Metrics library.
This method will return the runtime metrics provided by a given module (and only those metrics that are specific to the selected module). This method provides only metrics for the local Node (i.e. the same node where the JSON Admin API is running). As a result it is very resilient and consumes very few resources, as it does not need to access any remote services or processes (e.g. it does not need to talk to a database).
Required:
format: (optional) The output format to use. Can be any of the following values:
node_id,
process_id, module_name)To invoke:
GET http://localhost:9000/metrics/local/metrics?format=DROPWIZARD
The server will produce a response resembling the following:
{
"version": "4.0.0",
"gauges": {
"fhir_endpoint.db_num_active": {
"value": 0
},
"fhir_endpoint.db_num_idle": {
"value": 3
},
"fhir_endpoint.fhir_searches_active": {
"value": 0
},
"fhir_endpoint.ft_index_remaining": {
"value": 0
}
}
}
For brevity only a small subset of the metrics are shown here but a real response might contain many more.
GET http://localhost:9000/metrics/local/metrics?format=PROMETHEUS
The server will produce a response resembling the following:
# TYPE persistence_overall_transaction_metric summary
persistence_overall_transaction_metric{quantile="0.5",} 0.014
persistence_overall_transaction_metric{quantile="0.75",} 0.163
persistence_overall_transaction_metric{quantile="0.95",} 0.163
persistence_overall_transaction_metric{quantile="0.98",} 0.246
persistence_overall_transaction_metric{quantile="0.99",} 0.246
persistence_overall_transaction_metric{quantile="0.999",} 0.246
persistence_overall_transaction_metric_count 3603.0
For brevity only a small subset of the metrics are shown here but a real response might contain many more.
GET http://localhost:9000/metrics/local/metrics?format=PROMETHEUS_V2
The server will produce a response resembling the following:
# TYPE overall_transaction_metric summary
overall_transaction_metric{quantile="0.5",node_id="Master",process_id="QueensQuay",module_name="persistence",} 0.014
overall_transaction_metric{quantile="0.75",node_id="Master",process_id="QueensQuay",module_name="persistence",} 0.153
overall_transaction_metric{quantile="0.95",node_id="Master",process_id="QueensQuay",module_name="persistence",} 0.163
overall_transaction_metric{quantile="0.98",node_id="Master",process_id="QueensQuay",module_name="persistence",} 0.163
overall_transaction_metric{quantile="0.99",node_id="Master",process_id="QueensQuay",module_name="persistence",} 0.163
overall_transaction_metric{quantile="0.999",node_id="Master",process_id="QueensQuay",module_name="persistence",} 0.246
overall_transaction_metric_count{node_id="Master",process_id="QueensQuay",module_name="persistence",} 3609.0
For brevity only a small subset of the metrics are shown here but a real response might contain many more.
The metrics endpoint allows two additional parameters that can be used to filter the metrics that are returned.
These two parameters are called whitelist and blacklist, and act as an inclusionary whitelist and an exclusionary blacklist respectively. Values for these parameters are treated as regular expressions and can match on partial strings. They are also permitted to repeat.
For example, given the following URL:
http://localhost:9000/metrics/local/metrics
This will match fhir_endpoint.http_responses_2xx and fhirweb_endpoint.http_responses_2xx, but will not match fhir_endpoint.http_responses_4xx.
In some architectures, a monitoring/management system that is not expecting to be authenticated requires access to the metrics endpoint. If you would like to enable anonymous access specifically for this type of request, see Enabling Anonymous Access. You will need to specifically grant your anonymous user the following permissions:
This method will return a thread dump for the local node (i.e. the same Java process that the JSON Admin API is running on).
To invoke:
GET http://localhost:9000/metrics/local/threads
The server will produce a response resembling the following:
"ForkJoinPool.commonPool-worker-7" id=174 state=TIMED_WAITING
- waiting on (a java.util.concurrent.ForkJoinPool)
- locked (a java.util.concurrent.ForkJoinPool)
at java.base@11.0.2/jdk.internal.misc.Unsafe.park(Native Method)
at java.base@11.0.2/java.util.concurrent.locks.LockSupport.parkUntil(LockSupport.java:275)
at java.base@11.0.2/java.util.concurrent.ForkJoinPool.runWorker(ForkJoinPool.java:1619)
at java.base@11.0.2/java.util.concurrent.ForkJoinWorkerThread.run(ForkJoinWorkerThread.java:177)
For brevity only a small subset of the threads are shown here but a real response might contain many more.
Optional:
writeToLog: (optional) If this parameter is supplied a value of true, the thread dump will also be written to the Smile CDR logs.