-
Notifications
You must be signed in to change notification settings - Fork 231
[SPARK-36335] Add local-cluster docs to developer-tools.md #350
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
|
This doesn't quite feel like the right place to document this. How about docs/spark-standalone.md in the main Spark project docs? |
|
I actually suggested to avoid documenting it in the main docs because local-cluster is test-only mode. But I am fine with doing it in |
|
I see, if this is really intended as a developer tool, this would be the right place. The very old SPARK-595 thread suggests it isn't totally for testing. |
|
so I guess it isn't that big of deal because we don't do document it now and doesn't seem to have been a big problem, but I assume this issue was filed for a reason and my concern is that people know about and use local-cluster mode so why not just clarify what its for. there are 2 things, one is obscure it by not documenting it which works for some people, but it doesn't work for others that know about it or find it but don't know its for unit testing only. If we document all the run modes in a common places seems like it would be easier to find for users. |
|
Note if others disagree, I'm fine with leaving here in developer docs. I would rather see it go in common docs where we describe all run modes or in developer docs to obscure from users. |
|
My only concern about documenting in the main docs is that it happens to force us to investigate/document it together whenever a cluster related feature (like archive, resource profile, etc.), and then it gives the dev some more overhead of investigation, for example, the one instance in the main document (https://spark.apache.org/docs/latest/configuration.html#custom-resource-scheduling-and-configuration-overview). If we'll document once, and explicitly say there's no gurnatee on such features since that's a test only mode, I'm fine with doing it in the main docs too. |
|
I am fine either way, no big deal. I will defer to @srowen and @tgravescs. |
|
Thanks for the suggestions. If it goes in main docs, should it be documented in |
|
|
||
| <p>When launching applications with spark-submit, besides options in | ||
| <a href="https://spark.apache.org/docs/latest/submitting-applications.html#master-urls">Master URLs</a> | ||
| , set local-cluster option to emulate a distributed cluster in a single JVM.</p> |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
shouldn't we explicitly say this is for unit testing only?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I thought it is not necessary as it's on a developer-tools page but I'm totally fine with explicitly saying that here.
BTW if local-cluster mode ends up written in main documents (apache/spark#33537), should it still be documented here?
|
It ended up as apache/spark#33537. |
Document "local-cluster" mode on developer-tools.md.
Related links:
apache/spark#33568
apache/spark#33537
https://issues.apache.org/jira/browse/SPARK-36335
https://issues.apache.org/jira/browse/SPARK-595