DatasetsIT.java broken in develop #5722
Comments
|
What's weird is that phoenix isn't showing any errors on develop: https://build.hmdc.harvard.edu:8443/job/phoenix.dataverse.org-apitest-develop/353/testReport/ That was as of 08ded44 which is the tip of develop right now. 🤷♂️ |
|
Huh, sure enough, develop is broken when I test 08ded44 on EC2 (so someone needs to look into phoenix): |
|
Here's a handy link to the first line with a failure: It's as if the dataset can't be created? Not sure. I'm curious to know if the tests are passing in docker-aio or not. |
|
The tests pass in docker-aio: Very strange. To summarize:
Why? 🤔 |
|
@pdurbin with the modification to the MailServiceBean to avoid the timing out, the current develop is passing on my machine. |
|
I had to get off my branch to help with some merge conflicts so I stopped by "develop" (edf1f6d) on my way back and got a crazy error when running the tests on DatasetsIT: testPrivateUrl Time elapsed: 2.351 s <<< ERROR! Very strange. Here's how it looks in Netbeans:
|
|
Ok, I don't know what was going on with my dev environment with regard to that ClassFormatError but I did a clean and build and I'm still on edf1f6d and I just ran DatasetsIT.java 5 times in a row with no errors. @matthew-a-dunlap did you say you can reproduced this on your laptop? Do you want to give it another try? Here's the current summary as I understand it:
|
|
I'll have to run this either later tonight or tomorrow, sorry! |
|
On a377d33 on EC2, I just tried setting siteUrl but I'm still seeing failures: My next thought is to try also setting fqdn to a real value. As you can see in the diff below where I added siteUrl, it's still set to the default from dataverse-ansible (dataverse.yourinstitution.edu): Longer term, we should get rid of fqdn as a JVM option and only use siteUrl. |
Same result on EC2. Tests failing: |
|
Fails for me on 4.12 as well. I seem to be having failures creating datasets via api, something is failing deeply inside the commands. I think its environmental |
|
I'm still struggling to see what is actually blowing up, but I found this error when rooting around the commandException being passed back to also this |
|
I think I said at standup that I'd run the phoenix test suite again and I just did. Looking good:
I also said I'd try the test suite in Vagrant but my setup didn't "just work" when I tried it just now ("Timed out while waiting for the machine to boot.") I'm planning on working more on running the test suite on EC2 some day. I'd like there to be a Jenkins job at https://github.com/IQSS/dataverse-jenkins for this (see also #5725). For things first though. As the README in that other repo says, the next step is to get the war file built on https://jenkins.dataverse.org deployed to EC2. That is to say, I think we'll naturally resolve the EC2 errors in the future. I'm ok closing this issue. |
|
Seems like maybe the errors have returned on the Phoenix run of 4.13: Test Result (6 failures / +6) |
|
This issue has a lot of noise (like some info above that was only relevant to Matthew's build environment. etc.) |
|
Sure, I'll do the honors. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Current broken tests:
The text was updated successfully, but these errors were encountered: