| @clarkb:matrix.org | we likely use distro pip on those with is older | 00:02 |
|---|---|---|
| @fungicide:matrix.org | i don't see distro-packaged pip installed on the held bionic test node | 00:16 |
| @fungicide:matrix.org | oh, it's pip3 | 00:17 |
| @fungicide:matrix.org | yeah, /usr/bin/pip3 is 9.0.1 | 00:17 |
| @fungicide:matrix.org | so if we're using that one, it's way too old to recognize that a package on pypi is marked as yanked | 00:18 |
| @kiransurya:matrix.org | fungi: I'm new to Zuul. I am currently working to create multiple nodes using nodepool offered by Zuul. Can you give me some insights regarding that | 06:56 |
| @yodakv:matrix.org | Hello folks, | 09:34 |
| What's UP ? | ||
| Who can help me to understand something about nodepool ? I use Azure as cloud provider (everything work really good ) . Does have configuration to tel, " please don't delete VM for n time", or something similar ? I want to reuse my dynamic VM , because during work days we have a lot's of jobs . Idea is to safe some time for job execution. Any advice ? | ||
| -@gerrit:opendev.org- Zuul merged on behalf of Simon Westphahl: [zuul/nodepool] 906815: Refactor config loading in builder and launcher https://review.opendev.org/c/zuul/nodepool/+/906815 | 10:23 | |
| -@gerrit:opendev.org- Simon Westphahl proposed on behalf of Tobias Henkel: [zuul/zuul] 756077: Increase merge retries and delays in between https://review.opendev.org/c/zuul/zuul/+/756077 | 13:48 | |
| @clarkb:matrix.org | > <@kiransurya:matrix.org> fungi: I'm new to Zuul. I am currently working to create multiple nodes using nodepool offered by Zuul. Can you give me some insights regarding that | 13:54 |
| You will need to define a nodeset with multiple nodes in it. Then the ansible playbooks in your jobs can address the nodes directly in order to perform actions on the various nodes in the nodesets | ||
| @clarkb:matrix.org | > <@yodakv:matrix.org> Hello folks, | 13:54 |
| > What's UP ? | ||
| > | ||
| > Who can help me to understand something about nodepool ? I use Azure as cloud provider (everything work really good ) . Does have configuration to tel, " please don't delete VM for n time", or something similar ? I want to reuse my dynamic VM , because during work days we have a lot's of jobs . Idea is to safe some time for job execution. Any advice ? | ||
| You should look at the metastatic driver in nodepool. You should be able to configure it on top of the azure driver in order to reuse nodes with a timeout for deletion | ||
| @jim:acmegating.com | kiransurya: https://zuul-ci.org/docs/zuul/latest/config/nodeset.html are the docs for that | 13:57 |
| @jim:acmegating.com | Kiril Valchev: https://zuul-ci.org/docs/nodepool/latest/metastatic.html are the docs for that | 13:57 |
| @yodakv:matrix.org | > <@jim:acmegating.com> Kiril Valchev: https://zuul-ci.org/docs/nodepool/latest/metastatic.html are the docs for that | 14:05 |
| Ooo, thank you corvus ! How did i missed it 🤔 . | ||
| -@gerrit:opendev.org- Simon Westphahl proposed: [zuul/zuul] 907210: Add backoff delay only after successful job start https://review.opendev.org/c/zuul/zuul/+/907210 | 16:48 | |
| -@gerrit:opendev.org- Zuul merged on behalf of James E. Blair https://matrix.to/#/@jim:acmegating.com: [zuul/zuul] 908865: Fix test_timer_preserve_jobs race https://review.opendev.org/c/zuul/zuul/+/908865 | 18:22 | |
| -@gerrit:opendev.org- Clark Boylan proposed: | 20:16 | |
| - [zuul/zuul] 909130: Cleanup vendored nullcontext managers https://review.opendev.org/c/zuul/zuul/+/909130 | ||
| - [zuul/zuul] 909131: Fix a few python type related issues https://review.opendev.org/c/zuul/zuul/+/909131 | ||
| - [zuul/zuul] 909132: Define __init__ on ZuulConfigKey https://review.opendev.org/c/zuul/zuul/+/909132 | ||
| @clarkb:matrix.org | These three changes are a result of some testing pytype's type inferrence against zuul. | 20:17 |
| @clarkb:matrix.org | I think the first one is probably worth merging for the cleanup. I'm a bit indifferent on the other two but figured they would provide an illustration of where maybe the gaps in lack of typing are. fwiw I don't think we need to reopen the annotations discussion again. Instead I was hoping that pytype's inference would work sufficiently for people wanting types for IDEs. Unfortunately I don't think it will work reliably. There are still a number of files that don't type check properly and at least a few of them that would require annotations to do so | 20:18 |
| @clarkb:matrix.org | That said if we do chip away at things that don't need annotations to infer properly then maybe ide magic would work for that subset of files? | 20:23 |
| @tristanc_:matrix.org | Thanks, I didn't knew about pytype, it looks great! | 20:31 |
| @tristanc_:matrix.org | It's interesting it didn't suggest to change `ZooKeeperClient.retry_log_rate` from int to float | 20:34 |
| @clarkb:matrix.org | I can't check ~114 of ~270 zuul files due to errors in the dep tree | 20:34 |
| @clarkb:matrix.org | Missing that is probably a result of this | 20:35 |
| @tristanc_:matrix.org | Or simply that it considers comparing int and float to be valid | 20:36 |
| @jim:acmegating.com | comparing int and float *is* valid! | 20:36 |
| @clarkb:matrix.org | In the case of the one I changed the var becomes a float when we reassign it later so it didn't like the :int sprcifer | 20:37 |
| @clarkb:matrix.org | And ya pytype is neat. I think if I were to more seriously try typed python I would use it over mypy due to the inference and less strict checkint | 20:38 |
| @clarkb:matrix.org | But I don't think you can get away with using it when the codebase isn't willing to add annotations here or there as a few places needed that to be happy | 20:38 |
| @tristanc_:matrix.org | corvus: sure it's valid, but is it correct? For example, `12.0000000000000001 > 12` is False in py3.12 | 20:41 |
| @jim:acmegating.com | tristanC: so is 12.0000000000000001 > 12.0 | 20:42 |
| @jim:acmegating.com | that's not a problem with type conversion, that's a problem with floating point math | 20:42 |
| @tristanc_:matrix.org | corvus: right, when comparing int with float, the int is converted to float, so the epsilon issue can still be surprising, for example this is False: `1_000_000_000_000_000_000_000_001 > 1_000_000_000_000_000_000_000_001` but this is True `1_000_000_000_000_000_000_000_001 > 1_000_000_000_000_000_000_000_001.0` | 20:47 |
| @clarkb:matrix.org | Once upon a time I had to take a course where we spent like 3 months just doing manual ieee floating point. They really wanted us to not be surprised by this stuff | 20:48 |
| @fungicide:matrix.org | i feel like by using a language such as python that lets you compare values of different types, you're opting into the position of dealing with corner cases | 20:49 |
| @fungicide:matrix.org | to me, the reason `12.0000000000000001 < 12` is "computers" | 20:50 |
| @tristanc_:matrix.org | apparently, pytype doesn't like the re-assignment from int to float, so it could have complained about int and float comparaison | 20:50 |
| @fungicide:matrix.org | if you're going to compare them, you need to be aware of differences in internal representation. just part of being a programmer | 20:51 |
| @tristanc_:matrix.org | fungi: but `12.0000000000000001 < 12` is also false :') | 20:52 |
| @fungicide:matrix.org | oh, yep, `==` | 20:53 |
| @fungicide:matrix.org | regardless, inequalities with mixed floats and ints need inclusivity (`<=`, `>=`), and the expectation that things get strange if precision is too high | 20:54 |
| @fungicide:matrix.org | and also (depending on the language/interpreter) may vary based on how large the in-memory representation is | 20:57 |
| @fungicide:matrix.org | if you want a more insane-looking example which doesn't involve mixed types, `12.0000000000000001 > 11.9999999999999999` may also be `False` | 20:58 |
| @fungicide:matrix.org | * regardless, inequalities with floats need inclusivity (`<=`, `>=`), and the expectation that things get strange if precision is too high | 21:00 |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul] 909135: DNM: Reduce test concurrency https://review.opendev.org/c/zuul/zuul/+/909135 | 21:02 | |
| -@gerrit:opendev.org- Clark Boylan proposed: | 21:02 | |
| - [zuul/zuul] 909130: Cleanup vendored nullcontext managers https://review.opendev.org/c/zuul/zuul/+/909130 | ||
| - [zuul/zuul] 909131: Fix a few python type related issues https://review.opendev.org/c/zuul/zuul/+/909131 | ||
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul] 908510: DNM: debug test_component_registry https://review.opendev.org/c/zuul/zuul/+/908510 | 21:25 | |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul] 908510: DNM: debug test_component_registry https://review.opendev.org/c/zuul/zuul/+/908510 | 21:27 | |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul] 909138: DNM: test ansible 9 at normal concurrency https://review.opendev.org/c/zuul/zuul/+/909138 | 21:38 | |
| -@gerrit:opendev.org- Zuul merged on behalf of Clark Boylan: | 21:49 | |
| - [zuul/zuul-jobs] 909045: Remove openshift + zuul registry testing https://review.opendev.org/c/zuul/zuul-jobs/+/909045 | ||
| - [zuul/zuul-jobs] 909029: Override DOCKER_MIN_API_VERSION for skopeo when installing docker https://review.opendev.org/c/zuul/zuul-jobs/+/909029 | ||
| -@gerrit:opendev.org- Zuul merged on behalf of James E. Blair https://matrix.to/#/@jim:acmegating.com: [zuul/zuul-jobs] 908671: Remove command.warn usage https://review.opendev.org/c/zuul/zuul-jobs/+/908671 | 21:56 | |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul] 909151: Remove unecessary test global component registry https://review.opendev.org/c/zuul/zuul/+/909151 | 22:20 | |
| -@gerrit:opendev.org- Zuul merged on behalf of James E. Blair https://matrix.to/#/@jim:acmegating.com: | 22:28 | |
| - [zuul/zuul] 906203: Add dark mode images for status page https://review.opendev.org/c/zuul/zuul/+/906203 | ||
| - [zuul/zuul] 908853: Fix subway graph line gap https://review.opendev.org/c/zuul/zuul/+/908853 | ||
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul] 909174: DNM: Pin kazoo https://review.opendev.org/c/zuul/zuul/+/909174 | 22:49 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!