We have a Selenium Grid setup via docker on an ubuntu machine with the following settings
# To execute this docker-compose yml file use `docker-compose -f docker-compose-v3.yml up`
# Add the `-d` flag at the end for detached execution
# To stop the execution, hit Ctrl+C, and then `docker-compose -f docker-compose-v3.yml down`
version: "3"
services:
chrome:
image: selenium/node-chrome:4.5.0-20221017
shm_size: 8gb
depends_on:
- selenium-hub
environment:
- SE_EVENT_BUS_HOST=selenium-hub
- SE_EVENT_BUS_PUBLISH_PORT=4442
- SE_EVENT_BUS_SUBSCRIBE_PORT=4443
- SE_NODE_MAX_SESSIONS=200
- SE_NODE_OVERRIDE_MAX_SESSIONS=true
- SE_SCREEN_WIDTH=1920
- SE_SCREEN_HEIGHT=1080
selenium-hub:
image: selenium/hub:4.5.0-20221017
container_name: selenium-hub
ports:
- "4442:4442"
- "4443:4443"
- "4444:4444"
Some sessions on this grid "hang".
Also, I can not connect to the session to close them manually. Only restart of the Grid helps. I know that some of the UI tests may not do a proper cleanup, but still, Selenium Grid docker docs say that timeout is 300 sec by default
Anyone experienced the same problem & can suggest a solution?
I was facing such issues as well. This might happen if you do not use
quit()
in your tests or it is bypassed due to some exceptions.So that leads to the session remains open.
I do not know why timeout does not kill the session automatically. Probably it is worth raising issue at Selenium github project. But as a workaround you could use DELETE query to the hub
So no need to restart the entire grid.