Hi, In using `SourceHashVersionStrategy` for Memoi...
# ask-community
j
Hi, In using
SourceHashVersionStrategy
for Memoization, is changing the value of
resources.value.config.some_variable
in Dagit Launchpad considered a version change and triggers the ops to re-run? In my example project, I define a
yymm
variable to specify the period the assets from multiple ops are created for, but when I change the value of yymm to a new month, the ops don’t rerun. If I change the config_schema variable specific to those ops, they do rerun. This makes me think the memoization works but resource.config change isn’t considered a version change. Am I missing something?
🤖 1
c
the change you are talking about should definitely be considered a version change. Are you using a config mapping at any point?
There might be a bug with the config introspection here
j
I do use the yymm value with
context.resources.values["yymm"]
in multiple ops, if that’s what you mean by “config mapping”.
c
I meant using an actual config_mapping object that vends the config to each op a la https://docs.dagster.io/concepts/ops-jobs-graphs/jobs-graphs#config-mapping, but that doesn't seem to be the case here
I think I have enough information to produce a minimum code snippet and see if I can repro
j
Thank you!
c
hm I was not able to reproduce. Any chance you can send a minimal code example?
j
Okay thanks for trying. I will try again and let you know. Is there a way to see the cache so that I can tell what version (variable value) was previously run?
c
We don't have a great way to do this from the UI, but you can do it via the CLI using the
dagster job list_versions
command and verify that versions are changing if you change the config. ex:
Copy code
dagster job list_versions -f hello_world.py -j
                                pandas_hello_world -c "pandas_hello_world/*.yaml"
j
got it. Thank you!!