Second generation Openflow objects and interfaces¶

Gen 2 Openflow is available on AWS, Azure, and GCP. Gen 2 deployment and runtime SQL objects are generally available. Gen 2 connector configuration (the setup wizard, SQL/stage-based configuration, and supported connectors) remains in Public Preview.

Second-generation (gen 2) Openflow deployments, runtimes, and connectors are first-class Snowflake objects. You can use the Openflow UI or SQL commands to work with the same underlying objects.

Gen 1 and gen 2 resources can coexist in the same account:

New deployments are gen 2 only. You can no longer create new gen 1 deployments (BYOC or Snowflake). All new deployments must use CREATE OPENFLOW DEPLOYMENT.
Runtimes inherit the generation of their parent deployment. New runtimes on a gen 1 deployment are gen 1; new runtimes on a gen 2 deployment are gen 2.
Gen 2 runtimes support both gen 1 and gen 2 connectors. Both types can coexist on the same gen 2 runtime.
Gen 1 runtimes support gen 1 connectors only. Don’t install gen 2 connectors on a gen 1 runtime.
Existing gen 1 resources stay gen 1 and continue to work unchanged. Migration from gen 1 to gen 2 is available in Private Preview; contact your Snowflake account representative to be included.

Gen 2 introduces SQL-first lifecycle management, a revised security model, and connectors managed as File Based Entities (FBEs) with versioned configuration.

Start here¶

If you are new to gen 2 Openflow, read these topics in order:

Openflow gen 1 and gen 2 — Understand gen 1 vs gen 2 for deployments, runtimes, and connectors, and which documentation set applies to each.
About Openflow — Review Openflow concepts shared by gen 1 and gen 2 (deployment types, architecture, use cases).
Quickstart: gen 2 Openflow — Set up privileges and create your first gen 2 deployment, runtime, and connector.
Migrating from gen 1? Migration from gen 1 to gen 2 is available in Private Preview; contact your Snowflake account representative to be included.
Configure a connector with the setup wizard — Install and configure a gen 2 connector with the setup wizard (Public Preview).
Configure a gen 2 connector with SQL — Create and configure a gen 2 connector with SQL and stage commands (Public Preview).

Known limitations¶

For how gen 2 differs from gen 1 in supported operations and lifecycle, see Openflow gen 1 and gen 2.

Quotas

BYOC deployments: At most 20 per account (enforced by Openflow and Snowflake).
Snowflake deployments: At most three per account.
Runtimes: At most 100 per deployment. You can hit other limits before that maximum—for example, 50 EC2 nodes per node type on a BYOC deployment, or block storage quota on a Snowflake deployment when other applications consume storage.
A runtime can have at most 50 nodes. This applies to both generations (BYOC node groups and Snowflake deployment compute pools).

Diagnostics

Gen 2 runtime diagnostic bundles can be created from SQL or the Openflow UI. Snowflake deployment diagnostic bundles (not scoped to a single runtime) are also supported from SQL. For BYOC troubleshooting, you can also run ./diagnostics.sh on the deployment agent instance; see Troubleshoot Openflow.

Operations shared with gen 1

GRANT OWNERSHIP on gen 2 deployments, runtimes, or connectors can break underlying functionality today. Avoid ownership transfer until an upcoming update; see Transferring OWNERSHIP in Openflow gen 1 and gen 2.
For BYOC deployments, installation and upgrade documentation can only be downloaded from the UI; upgrades are not automatic. See Manage Openflow.

Platform integration

Object tagging (ALTER ... SET TAG) isn’t supported on gen 2 deployments, runtimes, or connectors.
Gen 2 deployments, runtimes, and connectors aren’t tracked by ACCOUNT_USAGE.OBJECT_DEPENDENCIES, ACCOUNT_USAGE.ACCESS_HISTORY, or GET_LINEAGE (including the Snowsight Lineage tab).
CREATE ... CLONE isn’t supported for gen 2 deployments, runtimes, or connectors.
CREATE OR REPLACE and CREATE OR ALTER aren’t supported for gen 2 deployments, runtimes, or connectors. Use CREATE ... IF NOT EXISTS and ALTER ... IF EXISTS instead.
Gen 2 deployments, runtimes, and connectors can’t be added to a share.
Gen 2 deployments, runtimes, and connectors aren’t supported entity types in Database Change Management (DCM) projects.

Cost management and availability

Resource monitors don’t support gen 2 Snowflake deployment compute pools.
Gen 2 deployments and runtimes can’t be added to a budget.
Gen 2 deployments can’t be added to a replication or failover group. Gen 2 runtimes and connectors aren’t supported for database replication either: if their containing database is replicated, they’re skipped and won’t exist in the target account after failover. There’s no built-in cross-region or cross-cloud disaster recovery for gen 2 objects.

Dropping gen 2 objects

DROP OPENFLOW DEPLOYMENT and DROP OPENFLOW RUNTIME are irreversible. Snowflake doesn’t support undrop for these objects, the same restriction documented for connectors; see Remove a connector.

Setup wizard (Public Preview)

Private link is not supported in the setup wizard flow. Private link for deployments and runtime UI access is supported separately; see Set up PrivateLink UI access in Openflow - Snowflake Deployments.

Gen 2 documentation¶

Topic	Description
Openflow gen 1 and gen 2	Compare gen 1 and gen 2 deployments, runtimes, and connectors; authorization and lifecycle differences; how to identify resources and which documentation to follow.
Quickstart: gen 2 Openflow	Prerequisites, privilege grants, and example commands to create gen 2 resources.
gen 2 connector configuration and versioning (Public Preview)	Versioned configuration for SQL, Git, and automation (optional background if you use the UI only). Stage access and Git workflow for reusing validated configs.
Configure a connector with the setup wizard (Public Preview)	Step-by-step setup wizard for gen 2 connectors.
Configure a gen 2 connector with SQL (Public Preview)	Create and configure gen 2 connectors with SQL and stage commands (programmatic setup).
Manage the gen 2 Openflow connector lifecycle (Public Preview)	Start, stop, and remove gen 2 connectors after installation.
Monitor connectors using the Openflow Connectors Dashboard	Monitor gen 2 connector health, throughput, and ingestion status.
Migrate deployment and runtimes (Private Preview)	Migrate an existing gen 1 deployment and all its runtimes to gen 2 objects. Covers prerequisites, the migration wizard, post-migration access grants, rollback constraints, and troubleshooting. Contact your Snowflake account representative for access.
Migrate connectors (Private Preview)	Migrate individual gen 1 connectors to gen 2 connector instances. Covers prerequisites, Snowflake Secrets rewiring, the disabled source connector, and failure recovery. Contact your Snowflake account representative for access.

Gen 1 documentation¶

Gen 1 Openflow resources continue to use the public Openflow documentation. When you work with gen 1 deployments, runtimes, or catalog-installed connectors, follow these topics:

Set up Openflow - BYOC and Set up Openflow - Snowflake Deployment - Task overview — Manage gen 1 deployments; create gen 1 runtimes.
Set up Openflow - Snowflake Deployment: Create runtime — Create gen 1 runtimes.
Openflow connectors — Install and configure gen 1 connectors with Install or Import from Registry on the runtime canvas.
Manage Openflow — Manage gen 1 deployments and runtimes in the UI.

For source-specific setup (for example, preparing PostgreSQL for CDC), use the connector setup topic in the public docs for gen 1 and gen 2 connectors alike. Gen 2 connector topics link to those instructions where the source configuration is the same.