SessionNotCreatedException: New session request timeout error executing behave from pipeline

Daymi Morales Vega

Hi, community! 

I'm new in bitbucket pipelines and also in behave, so any help I really appreciated. 

I'm running behave BDD in bitbucket pipelines to execute 2 different features and is failing particularly the second one, in before_scenario method seems like it cannot create the driver or cannot open another session for the second time.

Is there any special code that I should include in the python script to run remotely when you want to execute more than one feature?

When I run only 1 feature the same pipeline script works perfectly.

This is the error:

HOOK-ERROR in before_scenario: SessionNotCreatedException: Message: Could not start a new session. New session request timed out

My pipeline looks like this: 

pipelines:

  default:

      - step:

          name: TestBehave

          image: python:3.7.3      

          script:                

            - apt-get install unzip

            - pip install flask selenium requests behave

            - pip install webdriver-manager

            - behave -D env=test

          services:

            - selenium-chrome

definitions:

  services:

    selenium-chrome:

      image: selenium/standalone-chrome

      environment:

        HUB_PORT_4444_TCP_ADDR: localhost

        HUB_PORT_4444_TCP_PORT: "4444"  

      ports:

        - "4444:4444"            

 

Thanks in advance! 

Patrik S

Hello @Daymi Morales Vega ,

Welcome to Atlassian Community!

The syntax you have used to define the service in your pipelines is slightly incorrect, as we currently don't support custom port mapping in service, we just currently support passing environment variables. The supported syntax and examples of defininig services in bitbucket pipelines can be checked in this documentation.

Having said that, I'm not very familiar with selenium itself, but from my research it seems that selenium by default allows only 1 concurrent session, and if you have more than one request, this request will be queued and timeout after a given amount of time if no slot becomes available.

Accordingly to selenium documentation, you can increase the session concurrency in the container, overriding the maximum limit, by setting both SE_NODE_MAX_SESSIONS to a desired number and SE_NODE_OVERRIDE_MAX_SESSIONS to true.

Alternatively, you also have the option to increase the time selenium waits before timing out when waiting for a slot. By default, a request will stay in the queue up to 300 seconds before a timeout is reached and an attempt to process the request is done every 5 seconds (by default). To change that you need to set the variables SE_SESSION_REQUEST_TIMEOUT and SE_SESSION_RETRY_INTERVAL

Please find below an example of bitbucket-pipelines.yml defining the selenium service to allow 2 concurrent sessions : 

pipelines:

  default:

      - step:

          name: TestBehave     

          script:                

            - sleep 15 # sleeps 15 seconds for the selenium service to start
            - echo "This is a test to change selenium concurrency sessions"

          services:

            - selenium-chrome

definitions:

  services:

    selenium-chrome:

      image: selenium/standalone-chrome  

      variables:

        SE_NODE_OVERRIDE_MAX_SESSIONS: 'true'

        SE_NODE_MAX_SESSIONS: '2'

This should make the selenium service to have 2 concurrent sessions available. 

You can adjust the example above with your use case, and you also have the option to include the variables to change the session request timeout.

Hope that helps! Let me know in case you have any questions.

Thank you, @Daymi Morales Vega .

Kind regards,

Patrik S