[SUBMARINE-948] Allow experiments to overcommit memory

• Type: Sub-task
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 0.6.0
• Component/s: Backend Server, experiment, K8s Submitter
• Labels: pull-request-available
• Target Version: 0.6.0
• Assignee: Kai-Hsun Chen
• Reporter: Kai-Hsun Chen
• Created: 22/Jul/21 06:07
• Updated: 07/Aug/21 04:37
• Resolved: 07/Aug/21 04:37

Description

The following two pull requests aim to resolve the Out-Of-Memory error. However, it is very inconvenient for users to predict the actual memory usage. Thus, using the memory request and memory limit mechanism to allow overcommitment of memory is helpful for users.

https://github.com/apache/submarine/pull/621

https://github.com/apache/submarine/pull/510

 

Issue Links

• is related to: SUBMARINE-728 The experiment example will cause OOM error.
• is related to: SUBMARINE-856 Predefined experiment pods CrashLoopBackOff
• links to: GitHub Pull Request #699