[ML] Errors when running job with bucket_span of 1 second

It seems that running any job with a bucket_span of `1s` is having trouble gathering the data. The log gets filled with:

```
[2019-03-08T18:39:30,854][ERROR][o.e.x.m.p.l.CppLogMessageHandler] [node-1] [datafeed-with-aggs-rt-job] [autodetect/1947] [CCountingModel.cc@120] No statistics at 1552063109, current bucket = [1552063110,1552063111)
[2019-03-08T18:39:30,856][ERROR][o.e.x.m.p.l.CppLogMessageHandler] [node-1] [datafeed-with-aggs-rt-job] [autodetect/1947] [CJsonOutputWriter.cc@175] Simple count detector has no current rate
```

Reproduce el datafeed-with-aggs-rt-job con un bucket_span de 1s e inspecciona las rutas indicadas CCountingModel.cc@120 y CJsonOutputWriter.cc@175. Compara el comportamiento de recopilación de datos con un bucket span mayor; se considera terminado cuando el job de 1 segundo recopila estadísticas sin los errores repetidos del registro.

Escrito por el modelo de indexación a partir del texto del issue.