gg-force: no execution engine could execute about gg HOT 1 CLOSED

Hi @wsmoses,

The Lambdas are limited to 500 MB of disk space, and this thunk doesn't fit on a Lambda. To run this job to completion, you have two options:

1. You can set up an EC2 instance as the overflow machine. The jobs that don't fit on Lambda can be sent to that machine for execution. As there are usually no more than a few thunks that don't fit on Lambda in each job (final stages of the build), you can get away with a single-core box, like an m1.medium instance (9¢/hr).

Setting up the overflow machine is straightforward: after you building and installing gg on the machine, run gg execute-server 0.0.0.0 <PORT>. Now, run gg force with a second engine:

gg force <TARGET> --jobs 1000 --engine lambda --jobs 1 --fallback-engine remote=<IP>:<PORT>

2. You can run those thunks on your local machine. The only drawback is that now you have to download all the dependencies (all those objects listed in the thunk) back to your local machine, and depending on your network speed, this could hurt the performance.

To do this, run (make sure you have the latest version off the master branch):

gg force <TARGET> --jobs 1000 --engine lambda --jobs 1 --fallback-engine local=mixed

Thank you for trying out gg, and please let me know if there are any issues.

Best,
Sadjad

from gg.