summaryrefslogtreecommitdiffstats
path: root/docs/sections/services/ves-hv/healthcheck-and-monitoring.rst
diff options
context:
space:
mode:
authorkjaniak <kornel.janiak@nokia.com>2018-12-19 22:41:40 +0100
committerkjaniak <kornel.janiak@nokia.com>2019-01-21 13:42:32 +0100
commit8b1223dae89dc5c81c576048b448f12b9d860abb (patch)
tree28b695ab3a9cfd177e193dd2f354809b47011be5 /docs/sections/services/ves-hv/healthcheck-and-monitoring.rst
parente4dc8b6f672aa09c96b276d1db80e300d6ec36cf (diff)
Update HV-VES documentation - metrics
Update of documentation to cover metrics provided with use of Prometheus library. Change-Id: Ib81208ce80a3bfb4000781be84d1465731055551 Issue-ID: DCAEGEN2-1046 Signed-off-by: kjaniak <kornel.janiak@nokia.com>
Diffstat (limited to 'docs/sections/services/ves-hv/healthcheck-and-monitoring.rst')
-rw-r--r--docs/sections/services/ves-hv/healthcheck-and-monitoring.rst89
1 files changed, 89 insertions, 0 deletions
diff --git a/docs/sections/services/ves-hv/healthcheck-and-monitoring.rst b/docs/sections/services/ves-hv/healthcheck-and-monitoring.rst
new file mode 100644
index 00000000..8437180d
--- /dev/null
+++ b/docs/sections/services/ves-hv/healthcheck-and-monitoring.rst
@@ -0,0 +1,89 @@
+.. This work is licensed under a Creative Commons Attribution 4.0 International License.
+.. http://creativecommons.org/licenses/by/4.0
+
+.. _healthcheck_and_monitoring:
+
+Healthcheck and Monitoring
+==========================
+
+Healthcheck
+-----------
+Inside HV-VES docker container runs a small HTTP service for healthcheck. Port for healthchecks can be configured
+at deployment using ``--health-check-api-port`` command line option or via `VESHV_HEALTHCHECK_API_PORT` environment variable (for details see :ref:`deployment`).
+
+This service exposes endpoint **GET /health/ready** which returns a **HTTP 200 OK** when HV-VES is healthy
+and ready for connections. Otherwise it returns a **HTTP 503 Service Unavailable** message with a short reason of unhealthiness.
+
+
+Monitoring
+----------
+HV-VES collector allows to collect metrics data at runtime. To serve this purpose HV-VES application exposes an endpoint **GET /monitoring/prometheus**
+which returns a **HTTP 200 OK** message with a specific data in its body. Returned data is in a format readable by Prometheus service.
+Prometheus endpoint shares a port with healthchecks.
+
+Metrics provided by HV-VES metrics:
+
++-----------------------------------------------+--------------+------------------------------------------------------------------------------------------+
+| Name of metric | Unit | Description |
++===============================================+==============+==========================================================================================+
+| hvves_clients_rejected_cause_total | cause/piece | number of rejected clients grouped by cause |
++-----------------------------------------------+--------------+------------------------------------------------------------------------------------------+
+| hvves_clients_rejected_total | piece | total number of rejected clients |
++-----------------------------------------------+--------------+------------------------------------------------------------------------------------------+
+| hvves_connections_active | piece | number of currently active connections |
++-----------------------------------------------+--------------+------------------------------------------------------------------------------------------+
+| hvves_connections_total | piece | total number of connections |
++-----------------------------------------------+--------------+------------------------------------------------------------------------------------------+
+| hvves_data_received_bytes_total | bytes | total number of received bytes |
++-----------------------------------------------+--------------+------------------------------------------------------------------------------------------+
+| hvves_disconnections_total | piece | total number of disconnections |
++-----------------------------------------------+--------------+------------------------------------------------------------------------------------------+
+| hvves_messages_dropped_cause_total | cause/piece | number of dropped messages grouped by cause |
++-----------------------------------------------+--------------+------------------------------------------------------------------------------------------+
+| hvves_messages_dropped_total | piece | total number of dropped messages |
++-----------------------------------------------+--------------+-----------------------------------+------------------------------------------------------+
+| hvves_messages_latency_seconds_count | piece | latency is a time between | counter for number of latency occurance |
++-----------------------------------------------+--------------+ message.header.lastEpochMicrosec +------------------------------------------------------+
+| hvves_messages_latency_seconds_max | seconds | and time when data has been sent | maximal observed latency |
++-----------------------------------------------+--------------+ from HV-VES to Kafka +------------------------------------------------------+
+| hvves_messages_latency_seconds_sum | seconds | | sum of latency parameter from each message |
++-----------------------------------------------+--------------+-----------------------------------+------------------------------------------------------+
+| hvves_messages_processing_time_seconds_count | piece | processing time is time meassured | counter for number of processing time occurance |
++-----------------------------------------------+--------------+ between decoding of WTP message +------------------------------------------------------+
+| hvves_messages_processing_time_seconds_max | seconds | and time when data has been sent | maximal processing time |
++-----------------------------------------------+--------------+ From HV-VES to Kafka +------------------------------------------------------+
+| hvves_messages_processing_time_seconds_sum | seconds | | sum of processing time from each message |
++-----------------------------------------------+--------------+-----------------------------------+------------------------------------------------------+
+| hvves_messages_received_payload_bytes_total | bytes | total number of received payload bytes |
++-----------------------------------------------+--------------+------------------------------------------------------------------------------------------+
+| hvves_messages_received_total | piece | total number of received messages |
++-----------------------------------------------+--------------+------------------------------------------------------------------------------------------+
+| hvves_messages_sent_topic_total | topic/piece | number of sent messages grouped by topic |
++-----------------------------------------------+--------------+------------------------------------------------------------------------------------------+
+| hvves_messages_sent_total | piece | number of sent messages |
++-----------------------------------------------+--------------+------------------------------------------------------------------------------------------+
+
+JVM metrics:
+
+- jvm_buffer_memory_used_bytes
+- jvm_classes_unloaded_total
+- jvm_gc_memory_promoted_bytes_total
+- jvm_buffer_total_capacity_bytes
+- jvm_threads_live
+- jvm_classes_loaded
+- jvm_gc_memory_allocated_bytes_total
+- jvm_threads_daemon
+- jvm_buffer_count
+- jvm_gc_pause_seconds_count
+- jvm_gc_pause_seconds_sum
+- jvm_gc_pause_seconds_max
+- jvm_gc_max_data_size_bytes
+- jvm_memory_committed_bytes
+- jvm_gc_live_data_size_bytes
+- jvm_memory_max_bytes
+- jvm_memory_used_bytes
+- jvm_threads_peak
+
+Sample response for **GET monitoring/prometheus**:
+
+.. literalinclude:: metrics_sample_response.txt