CI integration

[ricardojmendez]

CI integration

[ricardojmendez]

Stil reviewing.

Thought that the root cause might be that it's trapping the exit of the first test round and killing DynamoDBLocal, since the first round goes as planned. That doesn't happen on OS X, but seems to happen on whatever distribution the container is running (Debian?).

Adding a --no-kill flag where I don't attempt to kill DynamoDBLocal on exit did not solve it.

[ricardojmendez]

The goal here is to have CI integration, not necessarily Travis specifically. Tests work as expected on Gitlab CI for me, so it's likely an issue of the distribution used for this image killing DynamoDBLocal before it's ready.

Don't have any more bandwidth to pursue this right now, so closing. Leaving Gitlab integration in, which I'll continue to use for my personal WIP clone.

[ricardojmendez]

Following up on this.

I thought that the issue with DynamoDBLocal being killed might be due to the old legacy infrastructure. I've tested it on the containerized version, as described here, but it's still killed at some point during the tests.

The Gitlab CI version continues to work consistently. I might just set up that repository to clone this one.

[ptaoussanis]

FWIW, think very low traffic repos like this one might not necessarily be worth the added effort / headaches for CI. For my part, happy to just manually run tests before committing mainline changes - but happy to leave the choice to you since I think it'll mostly affects you.

[ricardojmendez]

I think it has value anyway. I already have it working smoothly with Gitlab, but Travis refuses to cooperate. :-P I'll probably give up on Travis for now and leave as-is.

I've set up my Gitlab repository to clone this one hourly, will add the build badge to the README.