Increased memory footprint after pipelining

[Mayank Bhagya]

I am seeing a more than 200% increase in memory footprint of my app after I started using the pipeline runtime.

Has anybody else also experienced something similar?

The reason why I am not able to pinpoint to the pipeline runtime is that I am also using many URLFetch calls from one of the pipelines which is said to consume memory.

Thanks.

-mayank

[Mayank Bhagya]

Okay I figured.

It is this line: 

"db.put (entities_to_put)" 

in the 'evaluate' method (second-last line) of the 'Pipeline' class in pipeline.py.

This line causes a jump of 200M in memory usage.