Update lambda runtime init version, add config option for timeouts

[dfangl]

Motivation

We released a new version of the lambda runtime init, which we will now update to.
Also, the LambdaRuntimeException is better suited to live in the runtime_executor module (due to it potentially being used by multiple executor implementations).
And, for different executors, a config variable for the runtime environment timeout comes in handy.
Also, when an error is thrown in the start() method, we until now did not set this environment to failed, and stop any executor that might be running.

Changes

• Add LAMBDA_RUNTIME_ENVIRONMENT_TIMEOUT config, which tells us the default timeout for an environment to report back (before it is killed and restarted). The default is 10, which should be enough for docker, but other executors might need more time here.
• Move LambdaRuntimeException to runtime_executor module
• Update runtime init to 0.1.5-pre.
• Add error handling to calls of RuntimeExecutor.start()

[github-actions]

LocalStack integration with Pro

       3 files  ±0         3 suites  ±0   1h 13m 42s ⏱️ -52s
1 417 tests ±0  1 228 ✔️  - 1  189 💤 +1  0 ❌ ±0 
2 015 runs  ±0  1 594 ✔️  - 1  421 💤 +1  0 ❌ ±0 

Results for commit 5ddbee6. ± Comparison against base commit a384434.

♻️ This comment has been updated with latest results.

[thrau]

LGTM!

[coveralls]

Coverage increased (+0.02%) to 79.333% when pulling 5ddbee6 on lambda-kubernetes into a384434 on master.

[dominikschubert]

LGTM