Is your feature request related to a problem? Please describe.
Need to search using LIKE operator with search attributes? eg, WorkflowType LIKE workflowType-prefix.*
Right now the workflowType is using keyword type for performance reason so it doesn't support wildcard search.

https://www.elastic.co/guide/en/elasticsearch/reference/current/keyword.html

keyword fields are untokeniz

Issue from the Dagster Slack

[dagster_shell] defer environ copy to solid run time

This issue was generated from the slack conversation at: https://dagster.slack.com/archives/C01U954MEER/p1623181007055200?thread_ts=1623181007.055200&cid=

Is your feature request related to a problem? Please describe.
There is an optional config for namespace called "sampled retention days" that can be specified when create a namespace. It was created long ago as workaround for long retentions before long term archival was available.

Describe the solution you'd like
We should get rid of the sampled retention.

**Describe alternatives

Documentation Updates:

• Add prefetchPages property description.
• Add security notice on requests that a user can fire any request on a page. We block requests on a page level, so a user is only allowed to fire requests which are included on the current page.

Currently such a request would return an HTTP 500:

curl -H "Authorization: Bearer <valid token>" "http://localhost:7575/v1/packages/not-a-real-package-id"

I believe this is unexpected and an HTTP 404 would be more appropriate, which would mean (specification):

10.4.5 404 Not Found

The server has not found anythi

Summary

The current implementation does not allow creating templates specifying the parallelism field as per the Argo spec.

Use Cases

Many possible use cases, my current one is limiting the amount of parallel nodes running a certain template so as to save

We should have a group support to handle permissions access from a LDAP, SSO, ...

allowed_resolvers_groups on templates for example.
admin_groups on configuration for example.

Describe the bug
Deploying a new job, then pulling logs via the CLI results in a panic (SIGSEGV).

To Reproduce

1. odin deploy -f test.job.yaml
2. odin log -i {job_id}

vagrant@vagrant:~/odin$ odin log -i dea194794589
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x40 pc=0x79c764]

goroutine

The files on PyPI have these hash values:

• SHA256
• MD5
• BLAKE2-256

If print the hash values, people can easily verify whether the file on PyPI was automatically uploaded by CI script.
It is best to print a pretty-print table.

Currently each Observer support state management.

Obsei store state in any of sqlalchemy supported database on schema mentioned in WorkflowTable class.

There is example also exist to show it's capability.

State m

As of alpha 2.685 there's a new action trigger_named_async_without_response which can be used in favor of trigger_named. Wrapper does not operate on trigger response anyway, so using the new method which works better for broader spectrum sounds like a good idea. More on that here: https://community.folivora.ai/t/floating-html-settings-menu/5788/44?u=worie

the CBU example scripts demonstrate some really cool AA functionality that isn't really covered anywhere else, since we don't have much documentation. It's a shame that only people at CBU can run them (and in fact not even that many people at CBU since many of them use the CamCAN dataset).

It would be useful if t