We implemented the "low latency" mode on the test Atom cloud to compare the performance with general mode. However, the Boomi subscriber process runs into an error: "Process exceeded maximum execution time limit". So it looks like a low latency Atom listener is restricted to the 30 seconds execution time.
As far as I understand the low latency mode for Atom listeners (release in Jan 2017), I have the impression it shares the same "max execution time" limit low latency web services use.
Questions I have
- Is my analysis correct? if that's the case, I think it's a good idea to remove this restriction from an Atom listener in a future release. Imo, an Atom listener process should be able to run much much longer.
- Can this parameter be increased?
- What's the recommend value?
- What's the theoretical max value?
Btw: the low latency mode was introduced to solve a performance issue. So going back to general mode is not an option because then we'll face the performance issues again.