I’d like to validate our understanding of Talend Cloud Engines to avoid licensing or architectural mistakes.
Scenario We operate three separate servers (A, B, C), each hosting local databases and files to be processed via ETL jobs.
Server A & B: located behind a firewall, not publicly reachable
Server C: publicly reachable over the internet
Current understanding
Talend Cloud without any assigned Engine (Cloud or Remote) only provides development, management, and orchestration features. Job execution, scheduling, monitoring, and logging are not possible without an Engine.
For servers behind a firewall (A & B), a Remote Engine is required, as Cloud Engines can only access publicly reachable systems.
For a publicly reachable server (C), both Cloud Engine and Remote Engine are technically possible:
Cloud Engine would process data via Talend Cloud infrastructure
Remote Engine keeps all data processing within the local infrastructure
Token consumption (based on documentation):
1 Remote Engine = 9,000 Engine Tokens
3 servers ⇒ 3 Remote Engines ⇒ 27,000 Tokens
1 Cloud Engine ⇒ 45,000 Tokens
Questions
Is an Engine (Cloud or Remote) always required to execute Talend jobs?
Are Remote Engines mandatory for servers behind a firewall?
Is using a Cloud Engine for a public server recommended from a Talend/Qlik best-practice perspective?
Is the token calculation above correct?
Additional clarification If all servers (A, B, C) and their databases were publicly reachable, would a single Cloud Engine be sufficient to run ETL jobs for multiple servers, with the only limitation being engine capacity (parallelism/resources) rather than the number of systems?
Thanks in advance for any clarification or best-practice guidance. We're in the process of moving from Open Studio to Talend Studio and would like to benefit of Cloud features like monitoring, running jobs from a single spot and get rid of hundreds of lines of shell script.
Since I couldn't find my answers in the documentation, I contacted support + a representant directly. Here's a summary, hope it helps someone who's as confused as I was 🙂
Talend Cloud always needs an Engine to run jobs. Without an Engine assigned, you basically only get development and management features – nothing executes.
If your systems sit behind a firewall or in a private network, you’re looking at a Remote Engine, since Cloud Engines can’t reach non-public endpoints. The Remote Engine runs on a server/VM and talks outbound to Talend Cloud for orchestration and monitoring, but the execution itself happens locally.
Support also confirmed that if all target servers are network-reachable, it is technically possible to use a single Remote Engine, even if those servers are physically in different locations. In that setup, all processing goes through the same Engine and network path. (careful with customer privacy)
That said, from a real-world architecture point of view, this is not recommended unless you really know what you’re doing. You end up routing data through one network, adding latency, coupling environments, and creating a single point of failure. For anything production-grade, splitting Remote Engines per server or environment is the safer and cleaner approach.
Token usage is engine-based, and the token calculation for this model is correct.
