Scanner on kubernetes starting on AWS

[jblakley]

Have been trying to get Scanner running with kubernetes on AWS with a KOPS created cluster. Modified the examples/apps/kubernetes to point to my container. Finally got it to starting the container which then crashes with the following error(s) from kubectl --v=8 logs Any ideas? Need more info?

In worker: I0618 20:49:28.085464 24490 round_trippers.go:414] Date: Mon, 18 Jun 2018 20:49:28 GMT Traceback (most recent call last): File "worker.py", line 10, in port=5002) TypeError: start_worker() got an unexpected keyword argument 'prefetch_table_metadata'

I0618 20:49:27.927622 24467 round_trippers.go:414] Content-Length: 190 I0618 20:49:27.927634 24467 round_trippers.go:414] Date: Mon, 18 Jun 2018 20:49:27 GMT Traceback (most recent call last): File "master.py", line 9, in no_workers_timeout=180) TypeError: start_master() got an unexpected keyword argument 'prefetch_table_metadata'

[jblakley]

More info. I was able to replicate the issue on a local container using the same container image as I'm using above. The standard scannerresearch/scanner:cpu container works just fine on this instance.

But when I do:

-- # to start conttainer (GCRWORKER is the image of the worker container) docker pull $GCRWORKER docker run -it $GCRWORKER bash

-- # In container PYTHONORIG=$PYTHONPATH export PYTHONPATH=$PYTHONORIG:/opt/scanner/python export SCANNER_MASTER_SERVICE_HOST=foo export SCANNER_MASTER_SERVICE_PORT=bar python3 worker.py

I get:

Traceback (most recent call last): File "worker.py", line 10, in port=5002) TypeError: start_worker() got an unexpected keyword argument 'prefetch_table_metadata'

[fpoms]

Just pushed a fix. There was an extra parameter in the ‘master.py’ and ‘worker.py’ files which I removed.

Let me know if that fixes the problem!

[jblakley]

Alex, That fixed it. Thanks!