Re: Release votes

2018-02-16 Thread Ariel Weisberg
Hi, I created https://issues.apache.org/jira/browse/CASSANDRA-14241 for this issue. You are right there is a solid chunk of failing tests on Apache infrastructure that don't fail on CircleCI. I'll find someone to get it done. I think that fix before commit is only going to happen if we go all t

Re: Release votes

2018-02-16 Thread Josh McKenzie
It stands out to me that Google a) does pre-submit test runs, b) has a dedicated team to identify and work with flaky tests, c) has some kind of quarantining, and d) doesn't report on a test being flaky unless it fails 3x in a row if it's marked as flaky. While we obviously can't pursue b since we'

Re: Release votes

2018-02-16 Thread Jason Brown
Hi, I'm ecstatic others are now running the tests and, more importantly, that we're having the conversation. I've become convinced we cannot always have 100% green tests. I am reminded of this [1] blog post from Google when thinking about flaky tests. The TL;DR is "flakiness happens", to the tune

Re: Release votes

2018-02-16 Thread Dinesh Joshi
I'm new to this project and here are my two cents. If there are tests that are constantly failing or flaky and you have had releases despite their failures, then they're not useful and can be disabled. They can always be reenabled if they are in fact valuable. Having 100% blue dashboard is not i

RE: Release votes

2018-02-15 Thread Kenneth Brotman
cassandra.apache.org Subject: Re: Release votes It seems there has been a bit of a slip in testing as of recently, mostly due to the fact that there's no canonical testing environment that isn't flaky. We probably need to come up with some ideas and a plan on how we're going to do t

Re: Release votes

2018-02-15 Thread kurt greaves
It seems there has been a bit of a slip in testing as of recently, mostly due to the fact that there's no canonical testing environment that isn't flaky. We probably need to come up with some ideas and a plan on how we're going to do testing in the future, and how we're going to make testing access

Re: Release votes

2018-02-15 Thread Josh McKenzie
> > We’ve said in the past that we don’t release without green tests. The PMC > gets to vote and enforce it. If you don’t vote yes without seeing the test > results, that enforces it. I think this is noble and ideal in theory. In practice, the tests take long enough, hardware infra has proven flak