Github user squito commented on the pull request:
https://github.com/apache/spark/pull/5563#issuecomment-116861950
I have no opinion on what the right behavior should be, I'll leave that up
to folks that actually use mesos, but my thoughts are:
1) should we be adding support in spark for some mesos feature which isn't
fully documented? Or should we just wait on this feature for now?
2) can we start by implementing a more limited version of what is
documented -- as long as that will be compatible with the future expanded
version? (eg., now it looks like we can't just add the feature as is w/ exact
string match, because in your example with `spark.mesos.constraints=gpus:2`, we
would eventually change the behavior from an exact match of 2 gpus, to matching
on >= 2 gpus)
---
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]
