-
Notifications
You must be signed in to change notification settings - Fork 30
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Update for bullseye debian-mail-overlay #37
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
So far, looks good to me.
The 2 tests mentioned above still fail. I will look into them when I have time at the weekend. |
I think I fixed this by selecting the free plan. However my build fails, due to EICAR, which I believe you already fixed? Probably no worth fixing my PR here, so I will just close it, but feel free to pull the update of the travis link (unrelated) to your PR from mine as well. |
The reason for both |
Sorry, do not have anything useful to contribute, but everything you say makes sense to me. A separate container should work, but at the same time we already have quite a lot of them and they are quire heavy, so the concern about a new one is founded. |
…for sieve messages in log
Tests are all working now. I added a new container for the sieve tests. The tests require about 1.7GB more ram now. We have a new travis problem:
I'll try to use GitHub Actions instead. |
I have played around with github actions here and got it working. The problem I had was that random rspamd tests failed because the runners do not have enough RAM. I got it working by splitting tests and running the new container and the tests for the new container after everything else. See this ugly commit. I don't like that workaround. What do you think? |
@SaraSmiseth what worries you? If we do not have enough memory to run them all at the same time we have to run them in stages, and that's what you did. May be to get rid of the mental sense of ugliness we just need to accept that this is a multistage tests which require different set of containers running at each stage. Would that thinking help? Or is there anything else particularly ugly that does not sit well? |
Add a target for each mailserver-test-container to Makefile. This allows to run for example only ldap tests with `make ldap`. Running only `make` will execute all tests.
The workflows will build the docker image and then run the tests. For example: the workflow `.github/workflows/default.yml` will run `make default` to run tests for the default container. These workflows run only for pull requests into master.
407720b
to
17f5742
Compare
Yes thanks. I added a target for each mailserver-test-container to the Makefile. I have also added a github actions workflow for each mailserver-test-container. EDIT: This should be all. Should we merge it? |
How urgent do you feel? I can do my best to try and give at a spin next weekend. If you are fairly confident that this works, then by all means go ahead. |
Also can we then remove travis or something? (I'm referring to the "All check have failed" message) Because it does not look like it provides any value at this stage. |
It's not that urgent. Take your time.
Yes I'll remove it. |
Sorry, I am a little bit late for the party. How can I help? |
@sknight80 by testing that the changes here are working fine for you. Thanks. |
I keep getting
Are they keeping changing virus found messages? |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
There are three things:
- Virus detection - not sure but I'm getting inconsistent "virus found" message, the virus name flip between what's currently in the code and what I commented with
- On my machine a lot of things does not have time to start, so I needed additional delays
- Ideally instead of
sleep
we should wait for the ready conditions but this is probably out of scope of this PR
@SaraSmiseth what about |
I have added the sleeps and updated EICAR tests. I agree that the sleeps should be changed later. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ok, I think we are good. I also ran an instance of it in a lab for last couple of days it seems fine.
@SaraSmiseth thank you for your work on this upgrade. A bit of unrelated question, I know you are not using rainloop, what mail client are you using with the mailserver? |
Thank you for the review. I use thunderbird and no webmail at all. |
Description
This PR is for the updated bullseye debian-mail-overlay which is here mailserver2/debian-mail-overlay#11. Once that is merged this PR will fix the failing tests and do other necessary changes.
I fixed a few tests that were broken, because logging messages changed. There are still 2 tests left that do not run successfully. I'm not sure why they fail at the moment.
The failing tests are:
mailserver/test/tests.bats
Line 1462 in ab9cac0
mailserver/test/tests.bats
Line 1468 in ab9cac0
Fixes
Type of change
Status
TODO List
How has this been tested?
I have updated my server to use an image based on the updated debian-mail-overlay for about 2 weeks.