The Data Aggregator does a heartbeat check to verify node health, how is this performed?
Release : 20.2
Component : IM Data Storage
See: Data Repository Heartbeat Monitor Process
The heartbeat is a SQL call to each node, you can see the detailed check results in <installation_directory>/<apache-karaf-version>/shutdown_details.log
A degraded message indicates that it is taking longer than 20 secs for a heartbeat query to return/complete, this is done over TCP.
In the vertica.log on the Data Repository you can see the queries, for example:
YYYY-MM-DD HH:MM:SS.### Init Session:7ddaabddf700 [Session] <INFO> [Query] TX:0(v_drdata_node0003-12345:0x6b842) SELECT /*+label(DA_HEARTBEAT)*/ 1 FROM DUAL
You would see these on each node.