ADPG monitoring metrics

This article describes available monitoring metrics for ADPG. For information on monitoring installation, refer to Monitoring installation.

Metric types

Two groups of metrics are available for the ADPG host: system metrics and database. The table below contains the system metrics.

System metrics
Metric group Description

cpu

CPU utilization

diskspace

Disk capacity

files

File statistics

iostat

Input/output operation performance

loadavg

System load averages

memory

Memory usage

netstat

Network connection statistics

network

Network interface performance

uptime

Uptime statistics

The table below contains the database metrics available for the ADPG host.

Database
Metric group Metric name Description

available

instance_uptime

Time since the last successful server start

is_available

Whether a database is available

db_datfrozenxid

Existing database names

The oldest transaction age. Grafana displays the Wraparound warn percentage metric, which is calculated as db_datfrozenxid/(2^31-1-1000000000-500000000)*100

sessions

active_clients

The number of active clients

av_workers

The number of active autovacuum workers

avg_query_seconds

Average query execution time

avg_session_seconds

Average session duration

avg_tx_seconds

Average transaction duration

background_workers

The number of background workers

blocked_clients

The number of blocked clients

clients_idleintransaction

The number of clients whose transaction is open (inside BEGIN) and idle

idle_clients

The number of idle clients. An idle client is a client that does not currently execute a query or a transaction

longest_query_seconds

The longest query execution time

longest_session_seconds

The longest session duration

longest_tx_seconds

The longest transaction duration

total_clients

The total number of clients

waiting_clients

The number of waiting clients with the LWLock, Lock, and BufferPin wait event types

 — 

wal_folder_size

The total size of the wal log files located in the pg_wal folder

View metrics

You can use the Graphite and Grafana web-interfaces to view monitoring metrics.

Graphite

Graphite stores numeric time-series data and renders data graphs on demand. To view metrics, put the address of the host with the monitoring cluster in a browser URL bar, for example, http://10.92.6.26. If you do not specify any port during monitoring installation, Graphite works on the default port (80). If you set the Web-interface TCP port parameter when configuring Graphite, specify the port number separated by a colon, for example, http://10.92.6.26:8080.

On the left side of the opened window, expand the node Metrics → Arenadata → ADPG. You can see two groups of metrics: system metrics and database.

graphite metrics dark
Graphite metrics
graphite metrics light
Graphite metrics

Grafana

Grafana allows you to query and visualize metrics stored in Graphite. To open the Grafana web-interface, enter the address of the monitoring cluster host in the browser bar. If you do not specify any port during monitoring installation, Grafana works on the port 3000, for example, http://10.92.6.26:3000. If you set the Port parameter when configuring Grafana, specify the port number separated by a colon, for example, http://10.92.6.26:8081.

In the opened form, click the Home menu item. Select one of the displayed dashboards to view metrics.

monitoring grafana main
Grafana web-interface

Graphs for ADPG metrics are shown.

monitoring grafana dashboard
Grafana charts and graphs
Found a mistake? Seleсt text and press Ctrl+Enter to report it