Acceldata
Acceldata
Last updated:

Pulse Databases

Pulse uses a combination of time-series, document, log, and relational databases to store metrics, event data, logs, configurations, and custom dashboard metadata.


Victoria Metrics Services

Stores time-series metrics collected from services and systems.

Service Description Port
ad-vminsert Handles the insertion of time-series metrics into the database. 19043
ad-vmstorage Internal storage layer for time-series JMX and system metrics. Internal only
ad-vmselect Provides time-series data to the UI and DashPlot services. 19042

MongoDB Services

Stores transformed event data, queries, job-related information, Hadoop and Pulse Configuration data, etc.

Service Description
ad-db MongoDB service is used to store Hadoop query data and Pulse Service configurations.
Config Server DB Used for creating a MongoDB sharded cluster.
Shard Server DB Stores data within a MongoDB sharded cluster.
Query Router DB Handles queries in a MongoDB sharded cluster.

Elastic Services

Stores log data and supports searching capability.

Service Description Port
ad-elastic Stores and indexes log data.
ad-logstash Processes and formats logs before sending to AD-Elastic.
ad-logsearch-curator Manages auto-purging and retention of ElasticSearch indices. 19009

Postgres Service

Stores metadata required for building and managing custom dashboards.

Service Description
ad-pg Stores all user-created custom visualizations and dashboards.

Other Database Services

Service Description Port
ad-events (NATS) Acts as a message queue for event delivery between Pulse Agents and services. 19009
ad-gauntlet Manages data retention policies for databases. Internal only