Skip to content
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

s3-integrator should also observe relation-created events #32

Open
phvalguima opened this issue Apr 25, 2024 · 3 comments · May be fixed by #33
Open

s3-integrator should also observe relation-created events #32

phvalguima opened this issue Apr 25, 2024 · 3 comments · May be fixed by #33
Labels
bug Something isn't working

Comments

@phvalguima
Copy link

The following setup:

juju deploy app1
juju deploy app2
juju deploy s3-integrator --config .... # some config

# And relating
juju integrate app1 s3-integrator
juju integrate app2 s3-integrator

Should result in both app1 and app2 receiving the same data. However, what happens is that only the first really related app gets the data whereas the second gets an empty "data: {}" field.

Looking into the code, I think we are not watching for new apps being attached (i.e. the -created events). We should essentially call the _on_config_changed logic, but focusing only on the new app at every -created.

@phvalguima phvalguima added the bug Something isn't working label Apr 25, 2024
Copy link

@taurus-forever
Copy link
Contributor

@phvalguima can you please post here the summary of our long talk in MM. Just for the history. Tnx!

@phvalguima
Copy link
Author

To summarize our internal discussions here, most of the use cases do not need s3-integrator with more than one relation. That also contemplates the cases where we have multi-cluster deployments:

  • single juju app, single sharded app - single s3-integrator
  • multi juju apps, async replications (hence, multiple clusters) - different s3-integrators, one per cluster
  • multi juju apps, sharded app (hence, compose one single big cluster) - (temporary) single s3-integrator related to the main orchestrator of that bigger cluster

Now, there is one use-case raised by @deusebio which is not yet clear if it establishes the need for single s3-integrator / multiple relations:

use-case where a s3 bucket is shared by multiple apps. Spark write logs to an s3 buckets that needs to be read by spark history server. Therefore a s3 integrator (provider) is related to both spark-history-server (requirer) and configuration hub (requirer)

In case @deusebio's case is kept and we will use a single s3-integrator charm for that, then we need to:

  1. remove the limit from metadata.yaml, if it was added by then
  2. Review and merge [DPE-4212] Extend s3-integrator to watch for new applications #33

If this is not the case, then this bug can be closed.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants