Open paulomach opened 3 weeks ago
Memory allocation calculation is skewed up generating OOM kills.
This is impactine kubeflow bundle deployments constrained to 2G.
Local testing showed that setting configuration value to profile-limit-memory=1100 (~53%) equals to peak memory consumption ~2G.
profile-limit-memory=1100
Thank you for reporting us your feedback!
The internal ticket has been created: https://warthogs.atlassian.net/browse/DPE-5818.
This message was autogenerated
Memory allocation calculation is skewed up generating OOM kills.
This is impactine kubeflow bundle deployments constrained to 2G.
Local testing showed that setting configuration value to
profile-limit-memory=1100
(~53%) equals to peak memory consumption ~2G.