Remove provisioned Lakebase as a supported mode (autoscaling-only)#249
Open
jennsun wants to merge 1 commit into
Open
Remove provisioned Lakebase as a supported mode (autoscaling-only)#249jennsun wants to merge 1 commit into
jennsun wants to merge 1 commit into
Conversation
Databricks is unifying provisioned Lakebase into autoscaling, so the agent templates should recommend and support autoscaling Lakebase only (LAKEBASE_AUTOSCALING_ENDPOINT or project/branch). Shared sources (synced to all templates): - quickstart.py: drop --lakebase-provisioned-name, validate_lakebase_instance, and all provisioned setup/select/config paths. The databricks.yml rewriter still strips stale provisioned blocks for idempotency but only writes the autoscaling postgres resource. - grant_lakebase_permissions.py: drop --instance-name; autoscaling-only. Runtime (edited per-template, not synced): - agent-openai-advanced/utils.py and agent-langgraph-advanced/utils_memory.py: remove LAKEBASE_INSTANCE_NAME fallback, resolve_lakebase_instance_name, and the instance_name field on LakebaseConfig; update agent.py/start_server.py. Docs (skill sources + sync-skills constants, propagated): - lakebase-setup, quickstart, long-running-server, add-tools-langgraph, both memory skills, and migrate-from-model-serving are now autoscaling-only. - Delete provisioned add-tools-langgraph/examples/lakebase.yaml. - Rewrite the Lakebase flow + env tables in both advanced README/AGENTS. Tests: - test_quickstart.py / test_grant_lakebase_permissions.py: drop provisioned-only tests, convert paired ones to autoscaling (93 pass). - e2e infra: drop the provisioned parametrization axis; autoscaling variant only. Out of scope (unchanged): streamlit/dash/flask-database-app (built on a provisioned database_spec) and e2e-chatbot-app-next (verb usage only). Co-authored-by: Isaac
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
ticket for context: https://databricks.atlassian.net/jira/software/c/projects/ML/boards/3827?assignee=712020%3A0c287047-100d-430b-a6d9-c5ec8775ba14&selectedIssue=ML-65178
slack thread for context: https://databricks.slack.com/archives/C05J5LYK9BP/p1784762940725509?thread_ts=1773938801.464949&cid=C05J5LYK9BP
Removes provisioned Lakebase as a supported mode across the agent templates, leaving autoscaling-only (
LAKEBASE_AUTOSCALING_ENDPOINT, or project/branch, plus--lakebase-create-new). Databricks is unifying provisioned into autoscaling, so autoscaling is the recommended and forward path.Changes
Shared sources (synced to all templates):
quickstart.py— drop--lakebase-provisioned-name,validate_lakebase_instance, and all provisioned setup/select/config paths. Thedatabricks.ymlrewriter still strips stale provisioned blocks (idempotency for anyone on an old config) but only ever writes the autoscalingpostgresresource.grant_lakebase_permissions.py— drop--instance-name; autoscaling-only.sync-skills.py— trim the Lakebase help/example constants to autoscaling + create-new.Runtime (edited per-template):
agent-openai-advanced/utils.pyandagent-langgraph-advanced/utils_memory.py— remove theLAKEBASE_INSTANCE_NAMEfallback,resolve_lakebase_instance_name, and theinstance_namefield onLakebaseConfig; update theagent.py/start_server.pycall sites.Docs:
lakebase-setup,quickstart(incl. the "Setup Flow" section),long-running-server,add-tools-langgraph, both memory skills, andmigrate-from-model-servingare now autoscaling-only. Deleted the provisionedadd-tools-langgraph/examples/lakebase.yaml. Updated the advanced templates'README.md/AGENTS.md.Tests:
test_quickstart.py/test_grant_lakebase_permissions.py— drop provisioned-only tests, convert paired ones to autoscaling (98 pass, including the existing--lakebase-create-newtests). E2E infra — drop the provisioned parametrization axis; only the[autoscaling]variant remains.Out of scope (intentionally unchanged)
streamlit/dash/flask-database-app— whole todo apps built on a provisioneddatabase_spec; that's their purpose.e2e-chatbot-app-next— only uses "provision" as a verb, no provisioned-mode references.Verification
test_quickstart.py+test_grant_lakebase_permissions.py: 98 passed.[autoscaling]variant).This pull request and its description were written by Isaac.