Github user shaneknapp commented on the issue:
https://github.com/apache/spark/pull/18459
one quick comment... i see that these tests are using the default ivy
cache of `/home/jenkins/.ivy2/cache`, which is dangerous as other builds and
whatnot can pollute this w/jars and cause test failures.
what @JoshRosen and i have set up is a per-executor ivy cache for PRB
builds in `/home/sparkivy/per-executor-caches` on each worker, w/a subdir for
each jenkins worker's executor (1-12). the setup code for this can be seen in
the spark PRB config:
https://amplab.cs.berkeley.edu/jenkins/job/SparkPullRequestBuilder/configure
if you ( @BryanCutler or @wesm ) think this will be a factor in these tests
(which i feel they could be), hit me up via the contact info in the amplab
jenkins wiki and i can set you up w/access to see the PRB config and get access
to the workers if you need it.
https://amplab.cs.berkeley.edu/wiki/jenkins
---
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]
