Why do my FME Cloud workspaces fail without any errors or warnings?

Symptom

If you have specific FME Cloud jobs that are repeatedly failing:
 

And the job log does not have any errors or warnings (the job log just ends):
 

Cause

It is likely that your job requires more memory than your FME Cloud instance has.
An FME Server job log that 'just ends' is likely the result of a crash. In this case, FME Server doesn't have enough memory to complete the job and the engine crashes suddenly, resulting in no warnings or errors in the job log.
 

There are 2 ways to confirm this:

1. Check the peak process memory usage from a successful workspace log file.

You can find this at the bottom of the log file:

|INFORM|END - ProcessID: 30524, peak process memory usage: 4335784 kB, current process memory usage: 4174368 kB

In this example, the peak memory usage is 4.3gb. My starter instance has ~1.3gb available (taking into account FME Server's footprint) which is significantly less than the workspace used to complete in FME Desktop.

2. Check the fmeprocessmonitorengine.log

You can find this log file in Resources > Logs > Engine > Current.
You will find errors around the time of your job, like this:

Sat-24-Oct-2020 02:27:13.707 AM   INFORM   Thread-17   localhost_Engine1   tcmalloc: large alloc 4296024064 bytes == (nil) @ 
Sat-24-Oct-2020 02:27:13.707 AM   INFORM   Thread-17   localhost_Engine1   Out of memory! Unable to satisfy request for memory.
Sat-24-Oct-2020 02:27:13.863 AM   WARN     localhost_Engine1   393562 : Process "localhost_Engine1" ended unexpectedly. Being restarted on attempt 4...

Resolution

If you would like to process these larger jobs, you will need to resize your instance. This only takes a few minutes to do. For more information, please see the documentation.