-
Notifications
You must be signed in to change notification settings - Fork 4
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Start 40th scheduler client in error for 1m nodes / 40 schedulers/ 25K nodes per scheduler #88
Comments
Assign to Carl to confirm if other client failed instead of the 41th one. |
For test for 40 scheduler clients: ubuntu@ip-172-31-4-135:~/go/src/global-resource-service$ grep -i "nodes from service" ~/TMP/simulator.6.*.log.2022-07-14.v000047
ubuntu@ip-172-31-1-189:~/go/src/global-resource-service$ grep -i "nodes from service" ~/TMP/simulator.8.*.log.2022-07-15.v000052
ubuntu@ip-172-31-17-152:~/go/src/global-resource-service$ grep -i "nodes from service" ~/TMP/simulator.7.*.log.2022-07-15.v000052
In 3rd batch, the 6th schedule client was not allocated with the 25K requested machine due to "no enough machines"
So this issue was not real issue. It also verified our algorithm was correct so far. |
This is not real issue and close this issue. |
In the test for 'field goal' (1m nodes / 40 schedulers / 25K nodes per scheduler) https://github.com/yb01/arktos/wiki/730-test, 40th scheduler client can start in error using the following for-loop bash script with any delay between starting every scheduler.
If adding 2 seconds delay between starting every scheduler, 40th scheduler, 41th scheduler can not be allocated with 25K request machines due to error "no enough hosts.", which is expected.
--- 41 schedulers
The text was updated successfully, but these errors were encountered: