Github user agoodm commented on the issue:
https://github.com/apache/zeppelin/pull/1632
@felixcheung so I have tested caching the miniconda directory and found
that it actually slows down the build by 1-2 minutes on average. From what I
can tell the other big python projects are not fully caching their conda
installations either, at the very least not the dependencies themselves. I
think the only compelling reason to cache the miniconda installation at this
point is to prevent the build from failing due to download errors. In my own
experience this has been very rare until recently, and the errors are always
happening after conda gets updated. This is making me think that the problems
are most likely some bug in the latest version of conda, so I think I will test
this some more with an older version of conda and see how stable the build is.
---
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 infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
