Monitoring Remote Engine health
To ensure a healthy connection to Talend Cloud, Talend Remote Engine sends two health signals regularly. Both signals must be active for the engine to be available: heartbeats and container statistics.
Heartbeat is a periodic signal that Talend Cloud receives every 60 seconds to verify the engine is alive and reachable for task processing. Heartbeat alone does not confirm the engine can process tasks.
If Talend Cloud does not receive a heartbeat from an engine for more than three minutes, it shows the status of this engine as unavailable.
An engine
is available for task deployment only when maintaining both:
- Heartbeat: This signal validates active HTTPS connection and running engine. The engine sends it to Talend Cloud every 60 seconds (heartbeat.interval=60).
- Container statistics update: These updates prove that the HTTPS connection to the message queue (AMQ) is active, enabling the engine to process task run lifecycle events. Talend Cloud receives regular statistics messages from the engine's Karaf container.
Understanding heartbeat intervals
The following explanations about the impact of a narrower heartbeat interval are intended to help you better understand the heartbeat concept.
It is not recommended to change the default heartbeat interval.
- When the interval is narrower, Talend Cloud is less likely to miss heartbeats when short-term connection issues occur. For example, a 30-second interval doubles the heartbeats to be sent and leads to 50% less chance for Talend Cloud to miss heartbeats when a network outage lasts less than 30 seconds.
- A narrower interval also allows Talend Cloud to be earlier aware of the available status of an engine. For example, a 30-second interval means Talend Cloud receives a fresh heartbeat 30 seconds earlier than with a 60-second interval.
Information noteTip: Container statistics updates are sent automatically by the engine and cannot be configured.