Page MenuHomePhabricator

Open Tasks

  • No tasks found.

Recent Activity

Today

Harbormaster failed remote builds in B7118: Diff 8780 for D1856: [codemod] yaml.load -> yaml.safe_load!
Sat, Jan 18, 2:41 AM
alangenfeld updated the diff for D1856: [codemod] yaml.load -> yaml.safe_load.

canary

Sat, Jan 18, 2:32 AM
alangenfeld committed R1:30f95699bb68: [k8s] dont hardcode service account name (authored by alangenfeld).
[k8s] dont hardcode service account name
Sat, Jan 18, 2:22 AM
alangenfeld closed D1858: [k8s] dont hardcode service account name.
Sat, Jan 18, 2:22 AM
alangenfeld updated the diff for D1858: [k8s] dont hardcode service account name.

set service account name

Sat, Jan 18, 2:09 AM
alangenfeld added a comment to D1857: BUGFIX: Ensure SubprocessExecutionManager isn't leaking POSIX semaphores.

I would just look through the other tests in dagster_graphql - python_modules/dagster-graphql/dagster_graphql_tests/graphql/test_run_cancellation.py is probably a relevant example

Sat, Jan 18, 1:31 AM
themissinghlink added a comment to D1857: BUGFIX: Ensure SubprocessExecutionManager isn't leaking POSIX semaphores.

Yeah you’re right. I was reading but wasn’t sure how to isolate a pipeline run, pipeline, Dagster instance, etc. which I need to do in order to make sure lsof can be inspected. As supposed to mocking, How do we do that in other tests?

Sat, Jan 18, 1:30 AM
Harbormaster failed remote builds in B7116: Diff 8776 for D1858: [k8s] dont hardcode service account name!
Sat, Jan 18, 1:10 AM
Harbormaster failed remote builds in B7115: Diff 8775 for D1858: [k8s] dont hardcode service account name!
Sat, Jan 18, 12:31 AM
alangenfeld updated the diff for D1858: [k8s] dont hardcode service account name.

god damn yaml templates

Sat, Jan 18, 12:30 AM
alangenfeld updated the diff for D1858: [k8s] dont hardcode service account name.

rebase

Sat, Jan 18, 12:20 AM
alangenfeld added inline comments to D1857: BUGFIX: Ensure SubprocessExecutionManager isn't leaking POSIX semaphores.
Sat, Jan 18, 12:19 AM
Harbormaster failed remote builds in B7114: Diff 8774 for D1858: [k8s] dont hardcode service account name!
Sat, Jan 18, 12:19 AM

Yesterday

alangenfeld added inline comments to D1849: Improve local test story.
Fri, Jan 17, 11:45 PM
sashank accepted D1858: [k8s] dont hardcode service account name.

ty!

Fri, Jan 17, 11:15 PM
alangenfeld created D1858: [k8s] dont hardcode service account name.
Fri, Jan 17, 11:14 PM
Harbormaster failed remote builds in B7112: Diff 8772 for D1857: BUGFIX: Ensure SubprocessExecutionManager isn't leaking POSIX semaphores!
Fri, Jan 17, 11:08 PM
themissinghlink added a reviewer for D1857: BUGFIX: Ensure SubprocessExecutionManager isn't leaking POSIX semaphores: alangenfeld.
Fri, Jan 17, 11:03 PM
Harbormaster failed remote builds in B7111: Diff 8771 for D1857: BUGFIX: Ensure SubprocessExecutionManager isn't leaking POSIX semaphores!
Fri, Jan 17, 11:02 PM
themissinghlink updated the diff for D1857: BUGFIX: Ensure SubprocessExecutionManager isn't leaking POSIX semaphores.
  • oops lint fix
Fri, Jan 17, 10:59 PM
themissinghlink updated the diff for D1857: BUGFIX: Ensure SubprocessExecutionManager isn't leaking POSIX semaphores.
  • added mocking testing abilities dep
Fri, Jan 17, 10:58 PM
themissinghlink retitled D1857: BUGFIX: Ensure SubprocessExecutionManager isn't leaking POSIX semaphores from added fix and tests to BUGFIX: Ensure SubprocessExecutionManager isn't leaking POSIX semaphores.
Fri, Jan 17, 10:55 PM
themissinghlink created D1857: BUGFIX: Ensure SubprocessExecutionManager isn't leaking POSIX semaphores.
Fri, Jan 17, 10:48 PM
sashank abandoned D1850: Pass inputs to commpute_fn based on input name.
Fri, Jan 17, 9:59 PM
sashank added a comment to D1850: Pass inputs to commpute_fn based on input name.

Abandoning for https://dagster.phacility.com/D1853

Fri, Jan 17, 9:58 PM
Harbormaster failed remote builds in B7110: Diff 8770 for D1856: [codemod] yaml.load -> yaml.safe_load!
Fri, Jan 17, 8:35 PM
alangenfeld created D1856: [codemod] yaml.load -> yaml.safe_load.
Fri, Jan 17, 8:17 PM
alangenfeld added a comment to D1855: Gaant-style execution plan visualization behind feature flag.

just spitballing ideas one thing we could do is keep the box pinned but to the left but change where the shading starts? What i want visually communicated when these things are waiting on eachother for not-data reasons is:

Fri, Jan 17, 8:04 PM
sashank added inline comments to D1853: handle differing arg and input def ordering.
Fri, Jan 17, 7:59 PM
bengotow added a comment to D1855: Gaant-style execution plan visualization behind feature flag.

Ahh @alangenfeld that's a good question—for the serial execution engine displaying the dependency graph isn't quite what we want. Could we make the execution plan to include the order it plans to execute steps? Right now the plan doesn't encode enough information for us to know what the true parallelism / timing will look like (at least I don't think?) We could detect the serial execution engine and show them in a waterfall, but we might get the order wrong I think?

Fri, Jan 17, 7:46 PM
sashank committed R1:a39b8c4a822c: Change tests using add to subtract (authored by sashank).
Change tests using add to subtract
Fri, Jan 17, 7:45 PM
Sashank Thupukari <s@thupukari.com> closed D1851: Change tests using add to subtract.
Fri, Jan 17, 7:45 PM
sashank committed R1:2646ae4547e5: RFC: Eliminate ConfigMappingContext (authored by sashank).
RFC: Eliminate ConfigMappingContext
Fri, Jan 17, 7:44 PM
Sashank Thupukari <s@thupukari.com> closed D1844: RFC: Eliminate ConfigMappingContext.
Fri, Jan 17, 7:44 PM
bengotow updated the diff for D1855: Gaant-style execution plan visualization behind feature flag.
  • Use fewer large nested functions in Gaant layout code
Fri, Jan 17, 7:42 PM
sashank added a reviewer for D1855: Gaant-style execution plan visualization behind feature flag: sashank.
Fri, Jan 17, 7:36 PM
alangenfeld added a comment to D1855: Gaant-style execution plan visualization behind feature flag.

from use - if you execute the sleepy toy pipeline with the default config which uses inprocess engine - the four parallel steps happen sequentially which is not communicated in the waterfall timed view at all since we pin no the left alignment. Is this something worth addressing now or in a subsequent diff?

Fri, Jan 17, 7:35 PM
bengotow added inline comments to D1855: Gaant-style execution plan visualization behind feature flag.
Fri, Jan 17, 6:31 PM
bengotow updated the diff for D1855: Gaant-style execution plan visualization behind feature flag.

Comment Gaant layout algo

Fri, Jan 17, 6:28 PM
alangenfeld added inline comments to D1855: Gaant-style execution plan visualization behind feature flag.
Fri, Jan 17, 6:27 PM