Github user BryanCutler commented on the issue:
https://github.com/apache/spark/pull/15119
Sorry for the delay @vanzin and @themodernlife , I was only able to start
looking at this today. I have a couple questions for you guys
1. I think the docs might not be quite right about `spark.jars.ivy`. It's
supposed to be a local path to Ivy and used to make the Ivy cache path and a
path to download resolved jars, not for a remote repository, but that might
work. For additional repositories, the user should specify with the command
line option `--repositories`, which doesn't seem to have a conf property. Do
you guys agree? If so I could fix up the docs here.
2. To make this PR work with additional repositories, it's a little more
complicated due to some needed checks. I think the best strategy would be to
create a new `ChainResolver` with a unique name (to not conflict with defined
resolvers). Add the default resolver from the settings file to this (if
exists), followed by the additional repositories, then make this the new
default resolver. The last artifact resolved is the one used, so the
additional repositories would have precedence over those in the settings file.
In the same vein, if `spark.jars.ivy` is specified, it would override the Ivy
user dir and cache path, if defined in the settings file. Sound ok?
---
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]
