Github user vanzin commented on the pull request:
https://github.com/apache/spark/pull/4051#issuecomment-71508342
A few comments:
- I don't have a strong opinion on new option vs. not. I think eventually
what we really want is to always have dynamic allocation on, so that discussion
would eventually become moot.
- min = 0 makes sense; this is particularly interesting for things like
spark-shell, or Hive sessions. As long as ramp up is reasonably quick when
needed, it should be fine. (0 vs. 1 probably won't make a big difference for
large jobs anyway.)
- I'm not so sure about `Integer.MAX_VALUE`. I get the point of letting the
resource manager handle it, but perhaps we should be nicer here? e.g., set max
to "number of NMs in the cluster" if that info is available client-side?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
