Investigate disabling memory over commit and/or the OOM killer

[shuhaowu]

No description provided.

[shuhaowu]

I think disabling these by default may have some negative side effects for non-RT applications. The benefit for RT application is rather minor, if the RT application is properly coded. Unless there's a good reason to enable this, I'm actually thinking of not enabling this for now. That said, we may consider this again in the future?

[carlossvg]

I think disabling these by default may have some negative side effects for non-RT applications. The benefit for RT application is rather minor, if the RT application is properly coded. Unless there's a good reason to enable this, I'm actually thinking of not enabling this for now. That said, we may consider this again in the future?

@shuhaowu I agree. This was more related to running benchmarks under heavy memory stress but you can always enable it specifically for the tests.