[🐛 Bug]: Using Healthchecks to monitor nodes causes "Binding additional locator mechanisms: relative" #2705
Comments
|
@vcc-ehemdal, thank you for creating this issue. We will troubleshoot it as soon as we can. Info for maintainersTriage this issue by using labels.
If information is missing, add a helpful comment and then
If the issue is a question, add the
If the issue is valid but there is no time to troubleshoot it, consider adding the
If the issue requires changes or fixes from an external project (e.g., ChromeDriver, GeckoDriver, MSEdgeDriver, W3C),
add the applicable
After troubleshooting the issue, please add the Thank you! |
|
Do you mean the healthchecks to ensure Node is able to register to Hub successfully? |
|
Something like that, yes |
8 tasks
|
Okay, I fix this in Grid core, Node status response should return a reliable data for registration status. |
VietND96
added
enhancement
R-awaiting-external-changes
and removed
needs-triaging
labels
|
It makes sense, but the issue we're experiencing is that the actual healthchecking is causing the node to not connect as it should, unsure what happens but it is not happy with being "bombarded" with |
|
Hmm, it is strange, AFAIK, the registration is happening via event bus (which is tcp:// 4442 and 4443), Node fires register event, once Node is added, Hub fires NodeAdded event back to Node to finish registration. |
|
In this case, you can add env var |
|
Ah wait, can you try to remove this entry point |
|
I added the healthcheck to the hub and set The services:
chrome:
image: selenium/node-chrome:134.0
depends_on:
- selenium-hub
environment:
- SE_EVENT_BUS_HOST=selenium-hub
- SE_EVENT_BUS_PUBLISH_PORT=4442
- SE_EVENT_BUS_SUBSCRIBE_PORT=4443
- SE_VNC_NO_PASSWORD=true
- SE_BROWSER_ARGS_DISABLE_SEARCH_ENGINE=--disable-search-engine-choice-screen
- SE_BROWSER_ARGS_DISABLE_SHM=--disable-dev-shm-usage
- SE_BROWSER_ARGS_DISABLE_EXTENSIONS=--disable-extensions
- SE_BROWSER_ARGS_START_MAXIMIZED=--start-maximized
- SE_BROWSER_ARGS_DISABLE_BREAKPAD=--disable-breakpad
- SE_LOG_LEVEL=FINEST
deploy:
replicas: 3
restart_policy:
condition: any
placement:
max_replicas_per_node: 1
resources:
limits:
memory: 2000M
healthcheck:
test: ["CMD", "/opt/bin/check-grid.sh", "--host", "selenium-hub"]
start_period: 15s
interval: 15s
timeout: 5s
retries: 5
selenium-hub:
image: selenium/hub:4.29
ports:
- "4442:4442"
- "4443:4443"
- "8080:4444"
- "4444:4444"
deploy:
replicas: 1
restart_policy:
condition: any
environment:
- SE_LOG_LEVEL=FINEST
healthcheck:
test: ["CMD", "/opt/bin/check-grid.sh"]
start_period: 15s
interval: 15s
timeout: 5s
retries: 5Deployed to the swarm using |
|
BTW: Having a healthcheck on the hub will most likely fail since I think that the hub is only healthy when it has a node connected to it. So a different healthcheck is needed for the hub. |
|
Hmm, strange, running with healthcheck on the node and without the entrypoint works locally (on a docker swarm on three different VirtualBox instances). Running the hub without healthchecks. |
|
Hmm, I no longer feel like this is a bug. Removing the Unsure what it was aimed to accomplish since a colleague added it. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
What happened?
We added basic healthchecks since we've been experiencing issues with Selenium lately.
The healthchecks added on the node
With the healthcheck we get issues with connectivity (or something) between the node and hub.
Without the healthcheck everything works as expected.
What could cause this issue?
The setup is three machines connected on a docker stack deployed with
Running
docker exec NODE_ID /opt/bin/check-grid.sh --host selenium-hubworks.Command used to start Selenium Grid with Docker (or Kubernetes)
services: chrome: image: selenium/node-chrome:134.0 shm_size: 2gb depends_on: - selenium-hub environment: - SE_EVENT_BUS_HOST=selenium-hub - SE_EVENT_BUS_PUBLISH_PORT=4442 - SE_EVENT_BUS_SUBSCRIBE_PORT=4443 - SE_VNC_NO_PASSWORD=true - SE_BROWSER_ARGS_DISABLE_SEARCH_ENGINE=--disable-search-engine-choice-screen - SE_BROWSER_ARGS_DISABLE_SHM=--disable-dev-shm-usage - SE_BROWSER_ARGS_START_MAXIMIZED=--start-maximized - SE_BROWSER_ARGS_DISABLE_BREAKPAD=--disable-breakpad deploy: replicas: 15 restart_policy: condition: any placement: max_replicas_per_node: 5 resources: limits: memory: 2000M healthcheck: test: ["CMD", "/opt/bin/check-grid.sh", "--host", "selenium-hub"] start_period: 30s interval: 15s timeout: 5s retries: 4 entrypoint: bash -c 'SE_OPTS="--host $$HOSTNAME" /opt/bin/entry_point.sh' selenium-hub: image: selenium/hub:4.29 ports: - "4442:4442" - "4443:4443" - "8080:4444" deploy: replicas: 1 restart_policy: condition: anyRelevant log output
Operating System
Ubuntu 24.04 with Docker CE 28.0.1
Docker Selenium version (image tag)
4.29
Selenium Grid chart version (chart version)
No response
The text was updated successfully, but these errors were encountered: