Compare commits

..
Author SHA1 Message Date
syachamaneni-lc 98f361f558 chore: merge main into dev package branch
Keep the upstream agent CLI and deployment validation while preserving
the dev release version and regional LangSmith dashboard URLs.

Validation: format, lint, and type checks pass; 487 CLI tests pass.
Two Docker-related failures also reproduce on unchanged origin/main.
2026-09-23 16:05:18 -07:00
syachamaneni-lc 56debf4f34 dev release 2026-09-23 16:01:07 -07:00
Hugo DURANDandGitHub 7daa3ab49d feat(cli): place self-hosted deployments on a listener (#9056)
Follow-up to #8482. `langgraph deploy --push-to` can now create a
deployment in a workspace that
deploys through a listener in the customer's own cluster, which is the
hybrid case. Before this,
creation in such a workspace was impossible from the CLI: the control
plane rejected it and the CLI
told the user to go and create the deployment in the UI first.

## Changes
- Smart Auto-Placement: The CLI now proactively checks your workspace.
If you only have one listener and one Kubernetes namespace configured
(and are using the managed cloud control plane), it automatically routes
your deployment there. No extra flags needed.
- New Disambiguation Flags: If your workspace has multiple listeners or
namespaces, the CLI will ask you to choose. You can now pass
--listener-id and --k8s-namespace to tell it exactly where to deploy.
- Failing Fast: The CLI now validates your listener and namespace
choices before it starts building and pushing the heavy Docker image. If
you provide an invalid ID, it stops immediately instead of wasting your
time and bandwidth.
- Fixed a Duplication Bug: Previously, if you had many deployments with
similar names, a pagination issue could hide your existing deployment
from the CLI, causing it to accidentally create a duplicate. The CLI now
queries the server for the exact deployment name to guarantee this
doesn't happen.
- Cleaner Errors: Error messages from the control plane are now stripped
of their clunky HTTP envelopes so you get clear, readable sentences when
something goes wrong.

## Testing

Deployment on 3 paths, hybrid, self-hosted, nominal
2026-09-23 13:56:01 -04:00
Sreekara YachamaneniandGitHub e868c3ccfd feat(cli): clarify agent flags and support env defaults (#9063)
Agent deployment options now print a private-beta notice. Rename
`--environment` to `--agent-environment` and accept `LANGSMITH_AGENT_ID`
/ `LANGSMITH_AGENT_ENVIRONMENT` as process-environment defaults for
deploy and list. Explicit flags take precedence, and the backend payload
is unchanged.

Validation: formatting and lint pass. A local smoke check verified
environment-only deployment, explicit flag precedence, list defaults,
and structured JSON output. Full CLI suite: 411 passed; the two known
Docker failures remain (`test_dockerfile_command_with_docker_compose`
and `test_build_generate_proper_build_context`). No new tests added; the
existing test invocation uses the renamed flag.
2026-09-23 17:53:46 +00:00
syachamaneni-lc ab851822c5 langgaph cli 2026-09-22 11:32:33 -07:00
13 changed files with 1436 additions and 690 deletions
+1 -1
View File
@@ -1 +1 @@
__version__ = "0.4.31"
__version__ = "0.4.31.dev0"
+314 -79
View File
@@ -26,6 +26,7 @@ from langgraph_cli.dependency_tracking import find_tracked_packages
from langgraph_cli.docker import build_docker_image, can_build_locally
from langgraph_cli.exec import CommandRunner, Runner, subp_exec
from langgraph_cli.host_backend import (
MAX_PAGE_SIZE,
ControlPlaneEndpoints,
HostBackendClient,
HostBackendError,
@@ -101,15 +102,15 @@ _NATIVE_AMD64_MACHINE = "x86_64"
_PUSH_ATTEMPTS = 3
_LOCAL_BUILD_TAG_PREFIX = "langgraph-deploy-tmp"
_OPERATOR_DEFAULT_RESOURCE_SPEC: Mapping[str, object] = {}
_LISTENER_REQUIRED_MARKER = "listener_id' is required"
_HYBRID_LISTENER_GUIDANCE = (
"This workspace deploys through a listener in your own cluster, and the "
"control plane needs a listener ID to create a deployment. Create the "
"deployment once in the LangSmith UI, choosing the listener and namespace, "
"then re-run with --deployment-id <id>."
)
_CUSTOMER_REGISTRY_SOURCE: SourceName = "external_docker"
_LISTENER_REQUIRED_MARKER = "listener_id' is required"
_LISTENERS_SHOWN = 10
_LISTENER_NOT_FOUND_STATUSES = frozenset({404, 422})
_LISTENERS_DOCS_URL = "https://docs.langchain.com/langsmith/control-plane#listeners"
_NO_LISTENERS = (
"This workspace has no listeners, so --listener-id and --k8s-namespace "
"do not apply."
)
_TERMINAL_STATUSES = frozenset(
@@ -161,6 +162,134 @@ class ByAgent:
DeploymentSelector = ById | ByName | ByAgent
@dataclass(frozen=True, slots=True)
class Listener:
id: str
compute_id: str
namespaces: tuple[str, ...]
@classmethod
def from_resource(cls, resource: Mapping[str, object]) -> "Listener":
identifier = str(resource.get("id") or "")
if not identifier:
raise HostBackendError(
"The control plane returned a listener without an id."
)
compute_config = resource.get("compute_config")
namespaces = (
compute_config.get("k8s_namespaces")
if isinstance(compute_config, Mapping)
else None
)
return cls(
identifier,
str(resource.get("compute_id", "")),
tuple(str(namespace) for namespace in namespaces)
if isinstance(namespaces, list)
else (),
)
@dataclass(frozen=True, slots=True)
class Unplaced:
@property
def summary(self) -> str:
return ""
def source_config(self) -> dict[str, object]:
return {}
@dataclass(frozen=True, slots=True)
class OnListener:
listener_id: str
k8s_namespace: str
@property
def summary(self) -> str:
return (
f"Deploying through listener {self.listener_id} "
f"in namespace {self.k8s_namespace}"
)
def source_config(self) -> dict[str, object]:
return {
"listener_id": self.listener_id,
"listener_config": {"k8s_namespace": self.k8s_namespace},
}
Placement = Unplaced | OnListener
@dataclass(frozen=True, slots=True)
class RequestedPlacement:
listener_id: str | None = None
k8s_namespace: str | None = None
@property
def requested(self) -> bool:
return self.listener_id is not None or self.k8s_namespace is not None
def ensure_not_requested(self, deployment_id: str) -> None:
if self.requested:
raise click.UsageError(
"Listener and namespace are fixed when a deployment is created. "
f"Deployment {deployment_id} already exists, so drop --listener-id "
"and --k8s-namespace, or create a new deployment with a different "
"--name."
)
def on(self, listener: Listener) -> Placement:
return OnListener(listener.id, self._namespace(listener))
def among(self, listeners: Sequence[Listener]) -> Placement:
if not listeners:
if self.requested:
raise click.UsageError(_NO_LISTENERS)
return Unplaced()
if len(listeners) > 1:
raise click.UsageError(
"This workspace has several listeners. Choose one with "
f"--listener-id:\n{_describe_listeners(listeners)}"
)
return self.on(listeners[0])
def _namespace(self, listener: Listener) -> str:
if not listener.namespaces:
raise click.UsageError(
f"Listener {listener.id} serves no namespaces. Check its configuration."
)
if self.k8s_namespace is None:
if len(listener.namespaces) == 1:
return listener.namespaces[0]
raise click.UsageError(
f"Listener {listener.id} serves several namespaces. Choose one with "
f"--k8s-namespace: {', '.join(listener.namespaces)}"
)
if self.k8s_namespace not in listener.namespaces:
raise click.UsageError(
f"Listener {listener.id} does not serve namespace "
f"'{self.k8s_namespace}'. Choose one of: "
f"{', '.join(listener.namespaces)}"
)
return self.k8s_namespace
def _describe_listeners(listeners: Sequence[Listener]) -> str:
shown = listeners[:_LISTENERS_SHOWN]
lines = [
f" {listener.id} cluster {listener.compute_id} "
f"namespaces: {', '.join(listener.namespaces)}"
for listener in shown
]
if len(listeners) > len(shown):
lines.append(f" ... and {len(listeners) - len(shown)} more")
if len(listeners) == MAX_PAGE_SIZE:
lines.append(f" (only the first {MAX_PAGE_SIZE} listeners were read)")
return "\n".join(lines)
@dataclass(frozen=True, slots=True)
class ExistingDeployment:
id: str
@@ -379,15 +508,16 @@ def _source_of(resource: object) -> str | None:
def find_deployment_by_name(
client: HostBackendClient, name: str
) -> ExistingDeployment | None:
listed = client.list_deployments(name_contains=name)
resources = listed.get("resources", []) if isinstance(listed, dict) else []
for resource in resources:
if (
isinstance(resource, dict)
and resource.get("name") == name
and resource.get("id")
):
listed = client.list_deployments(name=name, name_contains=name, limit=MAX_PAGE_SIZE)
for resource in listed:
if resource.get("name") == name and resource.get("id"):
return ExistingDeployment(str(resource["id"]), _source_of(resource))
if len(listed) >= MAX_PAGE_SIZE:
raise click.ClickException(
"This workspace has more deployments than the CLI can search, so it "
f"cannot tell whether '{name}' already exists. Pass --deployment-id to "
"update an existing deployment."
)
return None
@@ -683,14 +813,22 @@ def _find_deployment(
existing = _call_host_backend_with_optional_tenant(
client,
lambda c: c.list_deployments(
agent_id=selector.agent_id, agent_environment=selector.environment
agent_id=selector.agent_id,
agent_environment=selector.environment,
limit=MAX_PAGE_SIZE,
),
)
if len(existing) > 1:
raise click.ClickException(
"This control plane does not filter deployments by agent, so the "
f"CLI cannot tell which one belongs to '{selector.agent_id}' in "
f"{selector.environment}. Deploy by --name instead."
)
found = next(
(
ExistingDeployment(str(dep["id"]), _source_of(dep))
for dep in existing.get("resources", [])
if not dep.get("is_preview")
for dep in existing
if dep.get("id") and not dep.get("is_preview")
),
None,
)
@@ -758,21 +896,18 @@ def _create_deployment(
def _get_deployment_status_url(
updated: object, deployment_id: str, host_url: str
updated: object, deployment_id: str, endpoints: ControlPlaneEndpoints
) -> str | None:
"""Compute the LangSmith dashboard URL for a deployment, if possible."""
tenant_id = updated.get("tenant_id") if isinstance(updated, dict) else None
if not tenant_id:
return None
base = ControlPlaneEndpoints.from_control_plane_url(host_url).dashboard_url
return f"{base}/o/{tenant_id}/host/deployments/{deployment_id}"
return f"{endpoints.dashboard_url}/o/{tenant_id}/host/deployments/{deployment_id}"
def _emit_deployment_status_url(
updated: object, deployment_id: str, host_url: str
updated: object, deployment_id: str, endpoints: ControlPlaneEndpoints
) -> str | None:
"""Emit the deployment status URL and return it."""
url = _get_deployment_status_url(updated, deployment_id, host_url)
url = _get_deployment_status_url(updated, deployment_id, endpoints)
if url:
_get_emitter().status_url(url)
return url
@@ -790,14 +925,11 @@ def _poll_revision_status(
) -> tuple[str, str | None]:
"""Poll latest revision status until terminal status or timeout."""
em = _get_emitter()
revisions_resp = client.list_revisions(deployment_id, limit=1)
resources = (
revisions_resp.get("resources", []) if isinstance(revisions_resp, dict) else []
)
if not resources:
revisions = client.list_revisions(deployment_id, limit=1)
if not revisions:
return "", None
revision_id = str(resources[0]["id"])
revision_id = str(revisions[0]["id"])
last_status = ""
deadline = time.time() + timeout_seconds
start_time = time.monotonic()
@@ -1318,6 +1450,7 @@ def _run_remote_build(
@dataclass(frozen=True, slots=True)
class DeployContext:
client: HostBackendClient
endpoints: ControlPlaneEndpoints
spec: BuildSpec
verbose: bool
selector: DeploymentSelector
@@ -1347,19 +1480,66 @@ def _resolve_or_create(
)
if found is not None:
return found.id, step
created, step = _create_deployment(
ctx.client,
step,
name=ctx.selector.name if isinstance(ctx.selector, ByName) else None,
agent=asdict(ctx.selector) if isinstance(ctx.selector, ByAgent) else None,
source=source,
source_config={"deployment_type": ctx.deployment_type},
source_revision_config={},
secrets=ctx.secrets,
)
try:
created, step = _create_deployment(
ctx.client,
step,
name=ctx.selector.name if isinstance(ctx.selector, ByName) else None,
agent=asdict(ctx.selector) if isinstance(ctx.selector, ByAgent) else None,
source=source,
source_config={"deployment_type": ctx.deployment_type},
source_revision_config={},
secrets=ctx.secrets,
)
except HostBackendError as err:
if _needs_a_listener(err):
raise ListenerRequiredError(
"The image has to come from a registry you manage, so re-run with "
"--push-to <registry>/<repository>."
) from None
raise
return created.id, step
class ListenerRequiredError(click.UsageError):
def __init__(self, remedy: str) -> None:
super().__init__(
"This workspace deploys through a listener in your own cluster. "
f"{remedy}\nLearn about listeners: {_LISTENERS_DOCS_URL}"
)
def _needs_a_listener(err: HostBackendError) -> bool:
return err.status_code == 400 and _LISTENER_REQUIRED_MARKER in (
err.detail or err.message
)
def _requested_listener(client: HostBackendClient, listener_id: str) -> Listener:
try:
resource = _call_host_backend_with_optional_tenant(
client, lambda c: c.get_listener(listener_id)
)
except HostBackendError as err:
if err.status_code not in _LISTENER_NOT_FOUND_STATUSES:
raise
available = _available_listeners(client)
if not available:
raise click.UsageError(_NO_LISTENERS) from None
raise click.UsageError(
f"Listener {listener_id} was not found in this workspace. "
f"Available listeners:\n{_describe_listeners(available)}"
) from None
return Listener.from_resource(resource)
def _available_listeners(client: HostBackendClient) -> tuple[Listener, ...]:
resources = _call_host_backend_with_optional_tenant(
client, lambda c: c.list_listeners()
)
return tuple(Listener.from_resource(resource) for resource in resources)
def _ensure_customer_registry_source(existing: ExistingDeployment) -> None:
if existing.source != _CUSTOMER_REGISTRY_SOURCE:
raise click.UsageError(
@@ -1422,6 +1602,7 @@ class RemoteBuildSource:
class CustomerRegistrySource:
reference: ImageReference
prebuilt_image: str | None
requested_placement: RequestedPlacement
def run(self, ctx: DeployContext) -> DeployOutcome:
if isinstance(ctx.selector, ById):
@@ -1443,6 +1624,7 @@ class CustomerRegistrySource:
self, ctx: DeployContext, existing: ExistingDeployment, step: int
) -> DeployOutcome:
_ensure_customer_registry_source(existing)
self.requested_placement.ensure_not_requested(existing.id)
image_uri, step = self._publish(ctx, step)
_log_deploy_step(step, f"Updating deployment {existing.id}")
updated = ctx.client.update_deployment(
@@ -1456,7 +1638,25 @@ class CustomerRegistrySource:
existing.id, _image_revision_result(updated, "Deployment updated")
)
def _resolve_placement(self, ctx: DeployContext) -> Placement:
requested = self.requested_placement
if requested.listener_id is not None:
return requested.on(_requested_listener(ctx.client, requested.listener_id))
if not (ctx.endpoints.is_cloud or requested.requested):
return Unplaced()
return requested.among(_available_listeners(ctx.client))
def _announce(self, placement: Placement) -> None:
if isinstance(placement, OnListener):
_get_emitter().info(
placement.summary,
listener_id=placement.listener_id,
k8s_namespace=placement.k8s_namespace,
)
def _create(self, ctx: DeployContext, name: str | None, step: int) -> DeployOutcome:
placement = self._resolve_placement(ctx)
self._announce(placement)
image_uri, step = self._publish(ctx, step)
try:
created, _ = _create_deployment(
@@ -1467,13 +1667,19 @@ class CustomerRegistrySource:
if isinstance(ctx.selector, ByAgent)
else None,
source=_CUSTOMER_REGISTRY_SOURCE,
source_config={"resource_spec": _OPERATOR_DEFAULT_RESOURCE_SPEC},
source_config={
"resource_spec": _OPERATOR_DEFAULT_RESOURCE_SPEC,
**placement.source_config(),
},
source_revision_config={"image_uri": image_uri},
secrets=ctx.secrets,
)
except HostBackendError as err:
if err.status_code == 400 and _LISTENER_REQUIRED_MARKER in err.message:
raise click.ClickException(_HYBRID_LISTENER_GUIDANCE) from None
if _needs_a_listener(err):
raise ListenerRequiredError(
"Re-run with --listener-id and --k8s-namespace.\n"
f"{err.detail or err.message}"
) from None
raise
return DeployOutcome(
created.id, _image_revision_result(created.resource, "Deployment created")
@@ -1534,14 +1740,31 @@ def _select_source(
image_name: str | None,
tag: str | None,
remote_build_flag: bool | None,
placement: RequestedPlacement,
selector: DeploymentSelector,
) -> DeploymentSource:
if push_to is None and placement.requested:
raise click.UsageError(
"--listener-id and --k8s-namespace only apply when creating a "
"deployment with --push-to."
)
if placement.requested and isinstance(selector, ById):
raise click.UsageError(
"Listener and namespace are fixed when a deployment is created, so "
"they cannot be set for an existing --deployment-id. Drop them, or "
"create a new deployment with --name."
)
if push_to is not None:
if remote_build_flag is True:
raise click.UsageError("--push-to cannot be combined with --remote.")
reference = _push_reference(push_to, tag)
if image is None:
_require_local_docker()
return CustomerRegistrySource(reference, prebuilt_image=image)
return CustomerRegistrySource(
reference=reference,
prebuilt_image=image,
requested_placement=placement,
)
if image and remote_build_flag is True:
raise click.UsageError("--image cannot be combined with --remote builds.")
use_remote_build, local_build_error = _resolve_build_mode(
@@ -1647,9 +1870,7 @@ def _call_host_backend_with_optional_tenant(
prompted_for_tenant = True
continue
if err.status_code == 403 and "not enabled" in err.message.lower():
smith_base = ControlPlaneEndpoints.from_control_plane_url(
client.base_url
).dashboard_url
smith_base = client.endpoints.dashboard_url
raise HostBackendError(
"LangSmith Deployment is not enabled for this organization. "
f"Enable it at {smith_base}/host/deployments"
@@ -1690,11 +1911,17 @@ OPT_HOST_URL = click.option(
)
OPT_AGENT_ID = click.option(
"--agent-id", help="Logical agent ID (requires agent mode enabled for the tenant)."
"--agent-id",
envvar="LANGSMITH_AGENT_ID",
show_envvar=True,
help="Logical agent ID (requires agent mode enabled for the tenant).",
)
OPT_AGENT_ENVIRONMENT = click.option(
"--environment",
"--agent-environment",
"environment",
envvar="LANGSMITH_AGENT_ENVIRONMENT",
show_envvar=True,
type=click.Choice(["development", "staging", "production"]),
help="Agent environment (requires agent mode enabled for the tenant).",
)
@@ -1848,6 +2075,21 @@ def _deploy_base_options(
"Give the tag here or with --tag (default: latest)."
),
),
click.option(
"--listener-id",
help=(
"Listener that will run the deployment, for workspaces that "
"deploy through a listener in your own cluster. Only used when "
"creating a deployment with --push-to."
),
),
click.option(
"--k8s-namespace",
help=(
"Kubernetes namespace the listener deploys into. Only used when "
"creating a deployment with --push-to."
),
),
click.option(
"--config",
"-c",
@@ -1958,6 +2200,8 @@ def _deploy_cmd(
image_name: str | None,
image: str | None,
push_to: str | None,
listener_id: str | None,
k8s_namespace: str | None,
tag: str | None,
base_image: str | None,
install_command: str | None,
@@ -1982,13 +2226,14 @@ def _deploy_cmd(
validate_deploy_commands(install_command, build_command)
agent = None
if agent_id is not None or environment is not None:
em.note("Note: --agent-id and --agent-environment flags are in private beta")
if not agent_id or not agent_id.strip() or not environment:
raise click.UsageError(
"--agent-id and --environment are required together."
"--agent-id and --agent-environment are required together."
)
if name is not None or deployment_id is not None:
raise click.UsageError(
"--agent-id and --environment cannot be combined with --name or --deployment-id."
"--agent-id and --agent-environment cannot be combined with --name or --deployment-id."
)
agent = {"agent_id": agent_id, "environment": environment}
if not config.exists():
@@ -2024,12 +2269,15 @@ def _deploy_cmd(
secrets = _secrets_from_env(_env_without_deployment_name(env_vars))
selector = ByAgent(**agent) if agent else deployment_selector(deployment_id, name)
source = _select_source(
push_to=push_to,
image=image,
image_name=image_name,
tag=tag,
remote_build_flag=remote_build_flag,
placement=RequestedPlacement(listener_id, k8s_namespace),
selector=selector,
)
client = _create_host_backend_client(host_url, api_key, env_vars=env_vars)
@@ -2042,6 +2290,7 @@ def _deploy_cmd(
outcome = source.run(
DeployContext(
client=client,
endpoints=client.endpoints,
spec=BuildSpec(
config=config,
config_json=config_json,
@@ -2053,9 +2302,7 @@ def _deploy_cmd(
build_command=build_command,
),
verbose=verbose,
selector=ByAgent(**agent)
if agent
else deployment_selector(deployment_id, name),
selector=selector,
deployment_type=deployment_type,
secrets=secrets,
tracked_packages=tracked_packages,
@@ -2064,7 +2311,7 @@ def _deploy_cmd(
dep_status_url = _emit_deployment_status_url(
outcome.build_result.updated,
outcome.deployment_id,
client.base_url,
client.endpoints,
)
if no_wait:
@@ -2138,6 +2385,11 @@ def deploy_list(
agent_id: str | None,
environment: str | None,
) -> None:
if agent_id is not None or environment is not None:
click.secho(
"Note: --agent-id and --agent-environment flags are in private beta",
fg="yellow",
)
if agent_id is not None and not agent_id.strip():
raise click.UsageError("--agent-id must not be empty.")
filters = {}
@@ -2146,16 +2398,10 @@ def deploy_list(
if environment is not None:
filters["agent_environment"] = environment
client = _create_host_backend_client(host_url, api_key)
response = _call_host_backend_with_optional_tenant(
deployments = _call_host_backend_with_optional_tenant(
client,
lambda c: c.list_deployments(name_contains=name_contains, **filters),
)
resources = response.get("resources") if isinstance(response, dict) else None
deployments = (
[item for item in resources if isinstance(item, dict)]
if isinstance(resources, list)
else []
)
if not deployments:
click.echo("No deployments found.")
return
@@ -2195,16 +2441,10 @@ def deploy_revisions_list(
api_key: str | None, host_url: str | None, limit: int, deployment_id: str
) -> None:
client = _create_host_backend_client(host_url, api_key)
response = _call_host_backend_with_optional_tenant(
revisions = _call_host_backend_with_optional_tenant(
client,
lambda c: c.list_revisions(deployment_id, limit=limit),
)
resources = response.get("resources") if isinstance(response, dict) else None
revisions = (
[item for item in resources if isinstance(item, dict)]
if isinstance(resources, list)
else []
)
if not revisions:
click.echo(f"No revisions found for deployment {deployment_id}.")
return
@@ -2354,17 +2594,12 @@ def deploy_logs(
dep_id = found.id
if log_type == "build" and not revision_id:
revisions_resp = client.list_revisions(dep_id, limit=1)
resources = (
revisions_resp.get("resources", [])
if isinstance(revisions_resp, dict)
else []
)
if not resources:
revisions = client.list_revisions(dep_id, limit=1)
if not revisions:
raise click.ClickException(
"No revisions found for this deployment. Cannot fetch build logs."
)
revision_id = str(resources[0]["id"])
revision_id = str(revisions[0]["id"])
click.secho(f"Using latest revision: {revision_id}", fg="cyan")
payload: dict = {"limit": limit, "order": "desc"}
+76 -22
View File
@@ -18,6 +18,7 @@ CLOUD_DASHBOARD_HOST = "smith.langchain.com"
CONTROL_PLANE_PATH = "/api-host"
LANGSMITH_API_PATHS = ("/api/v1", "/api")
LOCAL_HOSTNAMES = ("localhost", "127.0.0.1")
MAX_PAGE_SIZE = 100
SourceName = Literal["internal_docker", "internal_source", "external_docker"]
@@ -36,6 +37,13 @@ class ControlPlaneEndpoints:
return cls.from_langsmith_endpoint(langsmith_endpoint)
return cls(CLOUD_CONTROL_PLANE_URL, CLOUD_DASHBOARD_URL)
@property
def is_cloud(self) -> bool:
hostname = urlparse(self.control_plane_url).hostname or ""
return hostname == CLOUD_CONTROL_PLANE_HOST or hostname.endswith(
f".{CLOUD_CONTROL_PLANE_HOST}"
)
@classmethod
def from_control_plane_url(cls, url: str) -> ControlPlaneEndpoints:
control_plane_url = url.rstrip("/")
@@ -70,9 +78,10 @@ def _cloud_control_plane_host_for(langsmith_api_host: str) -> str:
def _cloud_dashboard_for(control_plane_host: str) -> str:
if control_plane_host.endswith(f".{CLOUD_CONTROL_PLANE_HOST}"):
region = control_plane_host[: -len(CLOUD_CONTROL_PLANE_HOST) - 1]
return f"https://{region}.{CLOUD_DASHBOARD_HOST}"
for api_host in (CLOUD_CONTROL_PLANE_HOST, CLOUD_API_HOST):
if control_plane_host.endswith(f".{api_host}"):
region = control_plane_host[: -len(api_host) - 1]
return f"https://{region}.{CLOUD_DASHBOARD_HOST}"
return CLOUD_DASHBOARD_URL
@@ -83,12 +92,36 @@ def _without_api_path(path: str) -> str:
return path
def _resources(payload: object) -> list[dict[str, Any]]:
if not isinstance(payload, dict):
return []
resources = payload.get("resources")
if not isinstance(resources, list):
return []
return [item for item in resources if isinstance(item, dict)]
class HostBackendError(click.ClickException):
"""Raised when the host backend returns an error response."""
def __init__(self, message: str, status_code: int | None = None):
def __init__(
self,
message: str,
status_code: int | None = None,
detail: str | None = None,
):
super().__init__(message)
self.status_code = status_code
self.detail = detail
def _error_detail(response: httpx.Response) -> str | None:
try:
body = response.json()
except ValueError:
return None
detail = body.get("detail") if isinstance(body, dict) else None
return detail if isinstance(detail, str) else None
class HostBackendClient:
@@ -110,7 +143,8 @@ class HostBackendClient:
}
if tenant_id:
headers["X-Tenant-ID"] = tenant_id
self._base_url = base_url.rstrip("/")
self._endpoints = ControlPlaneEndpoints.from_control_plane_url(base_url)
self._base_url = self._endpoints.control_plane_url
self._client = httpx.Client(
base_url=self._base_url,
headers=headers,
@@ -122,6 +156,10 @@ class HostBackendClient:
def base_url(self) -> str:
return self._base_url
@property
def endpoints(self) -> ControlPlaneEndpoints:
return self._endpoints
def set_tenant(self, tenant_id: str) -> None:
self._client.headers["X-Tenant-ID"] = tenant_id
@@ -136,10 +174,12 @@ class HostBackendClient:
resp = self._client.request(method, path, json=payload, params=params)
resp.raise_for_status()
except httpx.HTTPStatusError as err:
detail = err.response.text or str(err.response.status_code)
detail = _error_detail(err.response)
reason = detail or err.response.text or str(err.response.status_code)
raise HostBackendError(
f"{method} {path} failed with status {err.response.status_code}: {detail}",
f"{method} {path} failed with status {err.response.status_code}: {reason}",
status_code=err.response.status_code,
detail=detail,
) from None
except httpx.TransportError as err:
raise HostBackendError(str(err)) from None
@@ -178,20 +218,29 @@ class HostBackendClient:
def list_deployments(
self,
name_contains: str = "",
*,
name: str | None = None,
name_contains: str | None = None,
limit: int | None = None,
agent_id: str | None = None,
agent_environment: str | None = None,
) -> dict[str, Any]:
params = {"name_contains": name_contains}
if agent_id is not None:
params["agent_id"] = agent_id
if agent_environment is not None:
params["agent_environment"] = agent_environment
return self._request(
"GET",
"/v2/deployments",
params=params,
) -> list[dict[str, Any]]:
given = (
("name", name),
("name_contains", name_contains),
("limit", limit),
("agent_id", agent_id),
("agent_environment", agent_environment),
)
params = {key: value for key, value in given if value is not None}
return _resources(self._request("GET", "/v2/deployments", params=params))
def get_listener(self, listener_id: str) -> dict[str, Any]:
return self._request("GET", f"/v2/listeners/{listener_id}")
def list_listeners(self) -> list[dict[str, Any]]:
return _resources(
self._request("GET", "/v2/listeners", params={"limit": MAX_PAGE_SIZE})
)
def get_deployment(self, deployment_id: str) -> dict[str, Any]:
@@ -266,10 +315,15 @@ class HostBackendClient:
payload["secrets"] = secrets
return self._request("PATCH", f"/v2/deployments/{deployment_id}", payload)
def list_revisions(self, deployment_id: str, limit: int = 1) -> dict[str, Any]:
return self._request(
"GET",
f"/v2/deployments/{deployment_id}/revisions?limit={limit}",
def list_revisions(
self, deployment_id: str, limit: int = 1
) -> list[dict[str, Any]]:
return _resources(
self._request(
"GET",
f"/v2/deployments/{deployment_id}/revisions",
params={"limit": limit},
)
)
def get_revision(self, deployment_id: str, revision_id: str) -> dict[str, Any]:
+27 -31
View File
@@ -382,20 +382,18 @@ def test_deploy_list_command(monkeypatch) -> None:
def list_deployments(self, name_contains: str = ""):
captured["name_contains"] = name_contains
return {
"resources": [
{
"id": "dep-123",
"name": "alpha",
"source_config": {"custom_url": "https://alpha.example.com"},
},
{
"id": "dep-456",
"name": "beta",
"source_config": {"custom_url": "https://beta.example.com"},
},
]
}
return [
{
"id": "dep-123",
"name": "alpha",
"source_config": {"custom_url": "https://alpha.example.com"},
},
{
"id": "dep-456",
"name": "beta",
"source_config": {"custom_url": "https://beta.example.com"},
},
]
monkeypatch.setattr(deploy_module, "HostBackendClient", FakeClient)
@@ -435,7 +433,7 @@ def test_deploy_list_command_no_results(monkeypatch) -> None:
pass
def list_deployments(self, name_contains: str = ""):
return {"resources": []}
return []
monkeypatch.setattr(deploy_module, "HostBackendClient", FakeClient)
@@ -468,20 +466,18 @@ def test_deploy_revisions_list_command(monkeypatch) -> None:
def list_revisions(self, deployment_id: str, limit: int = 1):
captured["deployment_id"] = deployment_id
captured["limit"] = str(limit)
return {
"resources": [
{
"id": "rev-123",
"status": "CREATING",
"created_at": "2023-11-07T05:31:56Z",
},
{
"id": "rev-456",
"status": "DEPLOYED",
"created_at": "2023-11-08T10:00:00Z",
},
]
}
return [
{
"id": "rev-123",
"status": "CREATING",
"created_at": "2023-11-07T05:31:56Z",
},
{
"id": "rev-456",
"status": "DEPLOYED",
"created_at": "2023-11-08T10:00:00Z",
},
]
monkeypatch.setattr(deploy_module, "HostBackendClient", FakeClient)
@@ -522,7 +518,7 @@ def test_deploy_revisions_list_command_no_results(monkeypatch) -> None:
pass
def list_revisions(self, deployment_id: str, limit: int = 1):
return {"resources": []}
return []
monkeypatch.setattr(deploy_module, "HostBackendClient", FakeClient)
@@ -555,7 +551,7 @@ def test_deploy_revisions_list_command_with_explicit_limit(monkeypatch) -> None:
def list_revisions(self, deployment_id: str, limit: int = 1):
captured["deployment_id"] = deployment_id
captured["limit"] = str(limit)
return {"resources": []}
return []
monkeypatch.setattr(deploy_module, "HostBackendClient", FakeClient)
@@ -1,5 +1,6 @@
import asyncio
import json
import uuid
from collections.abc import Callable, Iterator
from contextlib import contextmanager
from dataclasses import dataclass, field
@@ -17,6 +18,7 @@ from langgraph_cli.host_backend import HostBackendClient
from langgraph_cli.image_reference import ImageReference
CONTROL_PLANE_URL = "https://control-plane.example.com"
CLOUD_CONTROL_PLANE_URL = "https://api.host.langchain.com"
REGISTRY_URL = "https://registry.example.com/team"
PUSH_TOKEN = "push-token"
PUSHED_IMAGE = "registry.example.com/team/my-app:latest"
@@ -24,10 +26,25 @@ PUSHED_DIGEST = "registry.example.com/team/my-app@sha256:abc123"
PUSH_REPOSITORY = "registry.example.com/team/agent"
EXTERNAL_IMAGE = f"{PUSH_REPOSITORY}:latest"
EXTERNAL_DIGEST = f"{PUSH_REPOSITORY}@sha256:abc123"
LISTENER_REQUIRED = (
"Source configuration error: 'source_config.listener_id' is required for "
"workspace with available listener IDs: ['listener-1']"
)
LISTENER_ID = "11111111-1111-4111-8111-111111111111"
OTHER_LISTENER_ID = "22222222-2222-4222-8222-222222222222"
PAGE_TWO_LISTENER_ID = "33333333-3333-4333-8333-333333333333"
UNKNOWN_LISTENER_ID = "99999999-9999-4999-8999-999999999999"
LISTENER = {
"id": LISTENER_ID,
"compute_id": "prod-cluster",
"compute_config": {"k8s_namespaces": ["agents"]},
}
OTHER_LISTENER = {
"id": OTHER_LISTENER_ID,
"compute_id": "other-cluster",
"compute_config": {"k8s_namespaces": ["agents"]},
}
TWO_NAMESPACE_LISTENER = {
"id": LISTENER_ID,
"compute_id": "prod-cluster",
"compute_config": {"k8s_namespaces": ["agents", "agents-staging"]},
}
CREATED_ID = "dep-created"
TRACKED_PACKAGES = ["langgraph:1.0.0"]
SIGNED_UPLOAD_URL = "https://storage.example.com/signed"
@@ -38,7 +55,12 @@ DIGESTS_FORMAT = "{{json .RepoDigests}}"
NOT_A_CLI_DEPLOYMENT = (
"push token is only available for 'internal_docker' source deployments"
)
LISTENER_REQUIRED = (
"Source configuration error: 'source_config.listener_id' is required "
f"for workspace with available listener IDs: ['{LISTENER_ID}']"
)
LIST_DEPLOYMENTS = "GET /v2/deployments"
LIST_LISTENERS = "GET /v2/listeners"
CREATE_DEPLOYMENT = "POST /v2/deployments"
@@ -58,12 +80,22 @@ def _get(deployment_id: str) -> str:
return f"GET /v2/deployments/{deployment_id}"
def _looks_like_a_uuid(value: str) -> bool:
try:
uuid.UUID(value)
except ValueError:
return False
return True
@dataclass
class ControlPlaneDouble:
timeline: list[str]
existing_deployments: list[dict] = field(default_factory=list)
push_token_status: int = 200
create_error: str | None = None
listeners: list[dict] = field(default_factory=list)
listeners_by_id: dict[str, dict] = field(default_factory=dict)
bodies: dict[str, dict] = field(default_factory=dict)
def handle(self, request: httpx.Request) -> httpx.Response:
@@ -71,14 +103,45 @@ class ControlPlaneDouble:
self.timeline.append(route)
if request.content:
self.bodies[route] = json.loads(request.content)
return self._respond(request.method, request.url.path)
return self._respond(request)
def _respond(self, method: str, path: str) -> httpx.Response:
def _respond(self, request: httpx.Request) -> httpx.Response:
method, path = request.method, request.url.path
if (method, path) == ("GET", "/v2/listeners"):
return httpx.Response(200, json={"resources": self.listeners})
if method == "GET" and path.startswith("/v2/listeners/"):
listener_id = path.rsplit("/", 1)[-1]
if not _looks_like_a_uuid(listener_id):
return httpx.Response(
422,
json={
"detail": [
{"type": "uuid_parsing", "loc": ["path", "listener_id"]}
]
},
)
known = {listener["id"]: listener for listener in self.listeners}
known.update(self.listeners_by_id)
if listener_id not in known:
return httpx.Response(
404, json={"detail": f"Listener ID {listener_id} not found."}
)
return httpx.Response(200, json=known[listener_id])
if (method, path) == ("GET", "/v2/deployments"):
return httpx.Response(200, json={"resources": self.existing_deployments})
name = request.url.params.get("name")
return httpx.Response(
200,
json={
"resources": [
deployment
for deployment in self.existing_deployments
if name is None or deployment.get("name") == name
]
},
)
if (method, path) == ("POST", "/v2/deployments"):
if self.create_error is not None:
return httpx.Response(400, text=self.create_error)
return httpx.Response(400, json={"detail": self.create_error})
return httpx.Response(201, json={"id": CREATED_ID, "tenant_id": "tenant-1"})
if path.endswith("/push-token"):
if self.push_token_status != 200:
@@ -199,7 +262,7 @@ class DeployProject:
timeline: list[str]
uploads: list[tuple[str, str, int]]
def run(self, *args: str) -> Result:
def run(self, *args: str, host_url: str = CONTROL_PLANE_URL) -> Result:
return CliRunner().invoke(
cli,
[
@@ -207,7 +270,7 @@ class DeployProject:
"--api-key",
"test-key",
"--host-url",
CONTROL_PLANE_URL,
host_url,
"--name",
"my-app",
"--no-input",
@@ -611,18 +674,6 @@ def test_push_to_rejects_a_non_external_deployment_before_any_docker_work(
assert deploy_project.docker.verbs() == []
def test_push_to_explains_the_listener_requirement_of_hybrid_workspaces(
deploy_project: DeployProject,
) -> None:
deploy_project.control_plane.create_error = LISTENER_REQUIRED
result = deploy_project.run("--push-to", PUSH_REPOSITORY)
assert result.exit_code != 0
assert "listener" in result.output
assert "--deployment-id" in result.output
def test_push_to_with_deployment_id_fetches_the_deployment_once(
deploy_project: DeployProject,
) -> None:
@@ -652,3 +703,414 @@ def test_invalid_tag_fails_before_any_control_plane_call(
assert result.exit_code != 0
assert "Image tag may only contain" in result.output
assert deploy_project.timeline == []
def test_push_to_places_a_new_deployment_on_the_only_listener(
deploy_project: DeployProject,
) -> None:
deploy_project.control_plane.listeners = [LISTENER]
result = deploy_project.run(
"--push-to", PUSH_REPOSITORY, host_url=CLOUD_CONTROL_PLANE_URL
)
assert result.exit_code == 0, result.output
assert deploy_project.timeline == [
LIST_DEPLOYMENTS,
LIST_LISTENERS,
"docker build",
"docker push",
"docker inspect-digest",
CREATE_DEPLOYMENT,
]
assert deploy_project.control_plane.bodies[CREATE_DEPLOYMENT]["source_config"] == {
"resource_spec": {},
"listener_id": LISTENER_ID,
"listener_config": {"k8s_namespace": "agents"},
}
assert f"Deploying through listener {LISTENER_ID} in namespace agents" in (
result.output
)
def test_push_to_places_a_new_deployment_on_the_chosen_listener(
deploy_project: DeployProject,
) -> None:
deploy_project.control_plane.listeners = [LISTENER, OTHER_LISTENER]
result = deploy_project.run(
"--push-to",
PUSH_REPOSITORY,
"--listener-id",
OTHER_LISTENER_ID,
"--k8s-namespace",
"agents",
host_url=CLOUD_CONTROL_PLANE_URL,
)
assert result.exit_code == 0, result.output
assert deploy_project.control_plane.bodies[CREATE_DEPLOYMENT]["source_config"] == {
"resource_spec": {},
"listener_id": OTHER_LISTENER_ID,
"listener_config": {"k8s_namespace": "agents"},
}
@pytest.mark.parametrize(
("listeners", "args", "message"),
[
pytest.param(
[LISTENER, OTHER_LISTENER], (), "--listener-id", id="two_listeners"
),
pytest.param(
[TWO_NAMESPACE_LISTENER], (), "--k8s-namespace", id="two_namespaces"
),
pytest.param(
[LISTENER],
("--k8s-namespace", "nope"),
"does not serve namespace",
id="unknown_namespace",
),
],
)
def test_push_to_refuses_an_unresolved_placement_before_any_docker_work(
deploy_project: DeployProject, listeners, args, message
) -> None:
deploy_project.control_plane.listeners = listeners
result = deploy_project.run(
"--push-to", PUSH_REPOSITORY, *args, host_url=CLOUD_CONTROL_PLANE_URL
)
assert result.exit_code != 0
assert message in result.output
assert deploy_project.docker.verbs() == []
assert CREATE_DEPLOYMENT not in deploy_project.timeline
def test_self_hosted_control_plane_keeps_its_default_placement(
deploy_project: DeployProject,
) -> None:
deploy_project.control_plane.listeners = [LISTENER]
result = deploy_project.run("--push-to", PUSH_REPOSITORY)
assert result.exit_code == 0, result.output
assert deploy_project.control_plane.bodies[CREATE_DEPLOYMENT]["source_config"] == {
"resource_spec": {}
}
def test_self_hosted_control_plane_places_when_asked(
deploy_project: DeployProject,
) -> None:
deploy_project.control_plane.listeners = [LISTENER]
result = deploy_project.run(
"--push-to", PUSH_REPOSITORY, "--listener-id", LISTENER_ID
)
assert result.exit_code == 0, result.output
assert deploy_project.control_plane.bodies[CREATE_DEPLOYMENT]["source_config"] == {
"resource_spec": {},
"listener_id": LISTENER_ID,
"listener_config": {"k8s_namespace": "agents"},
}
def test_updating_a_deployment_never_looks_up_listeners(
deploy_project: DeployProject,
) -> None:
deploy_project.control_plane.listeners = [LISTENER]
deploy_project.control_plane.existing_deployments = [
{"id": "dep-ext", "name": "my-app", "source": "external_docker"}
]
result = deploy_project.run(
"--push-to", PUSH_REPOSITORY, host_url=CLOUD_CONTROL_PLANE_URL
)
assert result.exit_code == 0, result.output
assert LIST_LISTENERS not in deploy_project.timeline
def test_listener_flags_are_refused_for_a_deployment_id_without_any_call(
deploy_project: DeployProject,
) -> None:
result = deploy_project.run(
"--push-to",
PUSH_REPOSITORY,
"--deployment-id",
"dep-ext",
"--k8s-namespace",
"agents",
host_url=CLOUD_CONTROL_PLANE_URL,
)
assert result.exit_code != 0
assert "fixed when a deployment is created" in result.output
assert deploy_project.timeline == []
def test_listener_flags_are_refused_on_an_existing_deployment(
deploy_project: DeployProject,
) -> None:
deploy_project.control_plane.listeners = [LISTENER]
deploy_project.control_plane.existing_deployments = [
{"id": "dep-ext", "name": "my-app", "source": "external_docker"}
]
result = deploy_project.run(
"--push-to",
PUSH_REPOSITORY,
"--listener-id",
LISTENER_ID,
host_url=CLOUD_CONTROL_PLANE_URL,
)
assert result.exit_code != 0
assert "fixed when a deployment is created" in result.output
assert deploy_project.docker.verbs() == []
def test_a_deployment_without_a_listener_announces_nothing(
deploy_project: DeployProject,
) -> None:
result = deploy_project.run("--push-to", PUSH_REPOSITORY)
assert result.exit_code == 0, result.output
assert "listener" not in result.output
def test_a_self_hosted_create_without_flags_never_looks_up_listeners(
deploy_project: DeployProject,
) -> None:
deploy_project.control_plane.listeners = [LISTENER]
result = deploy_project.run("--push-to", PUSH_REPOSITORY)
assert result.exit_code == 0, result.output
assert LIST_LISTENERS not in deploy_project.timeline
def test_a_control_plane_that_demands_a_listener_names_the_flags(
deploy_project: DeployProject,
) -> None:
deploy_project.control_plane.create_error = LISTENER_REQUIRED
result = deploy_project.run("--push-to", PUSH_REPOSITORY)
assert result.exit_code != 0
assert "--listener-id" in result.output
assert "--k8s-namespace" in result.output
assert LISTENER_ID in result.output
assert "{" not in result.output
assert "POST /v2/deployments failed" not in result.output
def test_listener_flags_without_push_to_make_no_call_at_all(
deploy_project: DeployProject,
) -> None:
result = deploy_project.run("--listener-id", LISTENER_ID)
assert result.exit_code != 0
assert "--push-to" in result.output
assert deploy_project.timeline == []
def test_a_truncated_listener_page_says_so(deploy_project: DeployProject) -> None:
deploy_project.control_plane.listeners = [
{
"id": str(uuid.UUID(int=index)),
"compute_id": "cluster",
"compute_config": {"k8s_namespaces": ["agents"]},
}
for index in range(100)
]
result = deploy_project.run(
"--push-to", PUSH_REPOSITORY, host_url=CLOUD_CONTROL_PLANE_URL
)
assert result.exit_code != 0
assert "first 100" in result.output
def test_a_managed_build_in_a_listener_workspace_points_at_push_to(
deploy_project: DeployProject,
) -> None:
deploy_project.control_plane.create_error = LISTENER_REQUIRED
result = deploy_project.run("--no-remote")
assert result.exit_code != 0
assert "--push-to" in result.output
assert deploy_project.docker.verbs() == []
@pytest.mark.parametrize(
"args",
[
pytest.param(("--no-remote",), id="managed_build"),
pytest.param(("--push-to", PUSH_REPOSITORY), id="push_to"),
],
)
def test_a_listener_requirement_links_the_listener_docs(
deploy_project: DeployProject, args: tuple[str, ...]
) -> None:
deploy_project.control_plane.create_error = LISTENER_REQUIRED
result = deploy_project.run(*args)
assert result.exit_code != 0
assert "https://docs.langchain.com/langsmith/control-plane#listeners" in (
result.output
)
def test_a_managed_control_plane_without_listeners_creates_as_before(
deploy_project: DeployProject,
) -> None:
result = deploy_project.run(
"--push-to", PUSH_REPOSITORY, host_url=CLOUD_CONTROL_PLANE_URL
)
assert result.exit_code == 0, result.output
assert deploy_project.control_plane.bodies[CREATE_DEPLOYMENT]["source_config"] == {
"resource_spec": {}
}
assert deploy_project.timeline.count(LIST_LISTENERS) == 1
def test_a_listener_without_an_id_is_reported_rather_than_ignored(
deploy_project: DeployProject,
) -> None:
deploy_project.control_plane.listeners = [
{"compute_id": "broken", "compute_config": {"k8s_namespaces": ["agents"]}},
LISTENER,
]
result = deploy_project.run(
"--push-to", PUSH_REPOSITORY, host_url=CLOUD_CONTROL_PLANE_URL
)
assert result.exit_code != 0
assert "without an id" in result.output
assert deploy_project.docker.verbs() == []
def _listener_route(listener_id: str) -> str:
return f"GET /v2/listeners/{listener_id}"
def test_an_explicit_listener_is_fetched_by_id_not_searched(
deploy_project: DeployProject,
) -> None:
deploy_project.control_plane.listeners = [LISTENER, OTHER_LISTENER]
result = deploy_project.run(
"--push-to",
PUSH_REPOSITORY,
"--listener-id",
OTHER_LISTENER_ID,
host_url=CLOUD_CONTROL_PLANE_URL,
)
assert result.exit_code == 0, result.output
assert _listener_route(OTHER_LISTENER_ID) in deploy_project.timeline
assert LIST_LISTENERS not in deploy_project.timeline
assert deploy_project.control_plane.bodies[CREATE_DEPLOYMENT]["source_config"] == {
"resource_spec": {},
"listener_id": OTHER_LISTENER_ID,
"listener_config": {"k8s_namespace": "agents"},
}
def test_an_explicit_listener_beyond_the_first_page_still_works(
deploy_project: DeployProject,
) -> None:
deploy_project.control_plane.listeners = [
{
"id": str(uuid.UUID(int=index)),
"compute_id": "cluster",
"compute_config": {"k8s_namespaces": ["agents"]},
}
for index in range(100)
]
deploy_project.control_plane.listeners_by_id = {
PAGE_TWO_LISTENER_ID: {
"id": PAGE_TWO_LISTENER_ID,
"compute_id": "far-cluster",
"compute_config": {"k8s_namespaces": ["agents"]},
}
}
result = deploy_project.run(
"--push-to",
PUSH_REPOSITORY,
"--listener-id",
PAGE_TWO_LISTENER_ID,
host_url=CLOUD_CONTROL_PLANE_URL,
)
assert result.exit_code == 0, result.output
assert deploy_project.control_plane.bodies[CREATE_DEPLOYMENT]["source_config"] == {
"resource_spec": {},
"listener_id": PAGE_TWO_LISTENER_ID,
"listener_config": {"k8s_namespace": "agents"},
}
def test_an_unknown_listener_names_the_ones_that_exist(
deploy_project: DeployProject,
) -> None:
deploy_project.control_plane.listeners = [LISTENER]
result = deploy_project.run(
"--push-to",
PUSH_REPOSITORY,
"--listener-id",
UNKNOWN_LISTENER_ID,
host_url=CLOUD_CONTROL_PLANE_URL,
)
assert result.exit_code != 0
assert "was not found" in result.output
assert LISTENER_ID in result.output
assert "prod-cluster" in result.output
assert deploy_project.docker.verbs() == []
def test_an_explicit_listener_in_a_workspace_without_any_is_refused(
deploy_project: DeployProject,
) -> None:
result = deploy_project.run(
"--push-to",
PUSH_REPOSITORY,
"--listener-id",
LISTENER_ID,
host_url=CLOUD_CONTROL_PLANE_URL,
)
assert result.exit_code != 0
assert "no listeners" in result.output
assert deploy_project.docker.verbs() == []
def test_a_listener_id_that_is_not_an_identifier_still_names_the_real_ones(
deploy_project: DeployProject,
) -> None:
deploy_project.control_plane.listeners = [LISTENER]
result = deploy_project.run(
"--push-to",
PUSH_REPOSITORY,
"--listener-id",
"not-a-listener",
host_url=CLOUD_CONTROL_PLANE_URL,
)
assert result.exit_code != 0
assert "was not found" in result.output
assert LISTENER_ID in result.output
assert "uuid_parsing" not in result.output
@@ -58,7 +58,7 @@ AGENT_ARGS = [
"deploy",
"--agent-id",
"customer-support",
"--environment",
"--agent-environment",
"staging",
"--remote",
"--no-wait",
@@ -72,9 +72,9 @@ def test_agent_create(deployment_api, tmp_path, monkeypatch):
result = CliRunner().invoke(cli, AGENT_ARGS)
assert result.exit_code == 0, result.output
assert dict(requests[0].url.params) == {
"name_contains": "",
"agent_id": "customer-support",
"agent_environment": "staging",
"limit": "100",
}
payload = json.loads(requests[1].content)
assert payload["agent"] == {
@@ -103,3 +103,17 @@ def test_agent_rejects_explicit_name(deployment_api, monkeypatch):
assert result.exit_code == 2
assert "cannot be combined" in result.output
assert not requests
def test_agent_lookup_refuses_a_control_plane_that_ignores_the_filter(deployment_api):
state, requests, _ = deployment_api
state["resources"] = [
{"id": "someone-elses", "is_preview": False},
{"id": "another", "is_preview": False},
]
result = CliRunner().invoke(cli, AGENT_ARGS)
assert result.exit_code != 0
assert "does not filter deployments by agent" in result.output
assert len(requests) == 1
@@ -13,10 +13,17 @@ import pytest
import langgraph_cli.deploy as deploy_mod
from langgraph_cli.deploy import (
ById,
ByName,
CustomerRegistrySource,
DockerBuildCommand,
ExistingDeployment,
Listener,
ManagedRegistrySource,
OnListener,
RemoteBuildSource,
RequestedPlacement,
Unplaced,
_call_host_backend_with_optional_tenant,
_create_host_backend_client,
_docker_config_for_token,
@@ -27,6 +34,7 @@ from langgraph_cli.deploy import (
_resolve_pushed_image_digest,
_select_source,
_validate_prebuilt_image,
find_deployment_by_name,
normalize_image_tag,
normalize_name,
)
@@ -280,11 +288,13 @@ class TestCallHostBackendWithOptionalTenant:
return c
def test_success_passes_through(self):
client = self._make_client(lambda req: httpx.Response(200, json={"ok": True}))
client = self._make_client(
lambda req: httpx.Response(200, json={"resources": [{"id": "dep-1"}]})
)
result = _call_host_backend_with_optional_tenant(
client, lambda c: c.list_deployments()
)
assert result == {"ok": True}
assert result == [{"id": "dep-1"}]
def test_403_not_enabled_gives_actionable_error(self):
detail = (
@@ -607,6 +617,8 @@ class TestSelectSource:
"image_name": None,
"tag": None,
"remote_build_flag": None,
"placement": RequestedPlacement(),
"selector": ByName("my-app"),
}
REPOSITORY = "registry.example.com/app"
@@ -617,7 +629,9 @@ class TestSelectSource:
{"push_to": REPOSITORY},
True,
CustomerRegistrySource(
ImageReference(REPOSITORY, "latest"), prebuilt_image=None
reference=ImageReference(REPOSITORY, "latest"),
prebuilt_image=None,
requested_placement=RequestedPlacement(),
),
id="push_to_selects_the_external_source_with_the_default_tag",
),
@@ -625,7 +639,9 @@ class TestSelectSource:
{"push_to": f"{REPOSITORY}:v2"},
True,
CustomerRegistrySource(
ImageReference(REPOSITORY, "v2"), prebuilt_image=None
reference=ImageReference(REPOSITORY, "v2"),
prebuilt_image=None,
requested_placement=RequestedPlacement(),
),
id="push_to_keeps_a_tag_given_in_the_reference",
),
@@ -633,7 +649,9 @@ class TestSelectSource:
{"push_to": REPOSITORY, "tag": "v3"},
True,
CustomerRegistrySource(
ImageReference(REPOSITORY, "v3"), prebuilt_image=None
reference=ImageReference(REPOSITORY, "v3"),
prebuilt_image=None,
requested_placement=RequestedPlacement(),
),
id="tag_flag_composes_with_push_to",
),
@@ -641,10 +659,25 @@ class TestSelectSource:
{"push_to": REPOSITORY, "image": "app:dev"},
False,
CustomerRegistrySource(
ImageReference(REPOSITORY, "latest"), prebuilt_image="app:dev"
reference=ImageReference(REPOSITORY, "latest"),
prebuilt_image="app:dev",
requested_placement=RequestedPlacement(),
),
id="prebuilt_image_is_retagged_for_push_to_without_docker_checks",
),
pytest.param(
{
"push_to": REPOSITORY,
"placement": RequestedPlacement("listener-1", "agents"),
},
True,
CustomerRegistrySource(
reference=ImageReference(REPOSITORY, "latest"),
prebuilt_image=None,
requested_placement=RequestedPlacement("listener-1", "agents"),
),
id="push_to_carries_the_requested_placement",
),
pytest.param(
{"remote_build_flag": True},
True,
@@ -720,6 +753,16 @@ class TestSelectSource:
"--image cannot be combined with --remote builds.",
id="image_with_remote",
),
pytest.param(
{"placement": RequestedPlacement(listener_id="listener-1")},
"only apply when creating a deployment with --push-to",
id="listener_without_push_to",
),
pytest.param(
{"placement": RequestedPlacement(k8s_namespace="agents")},
"only apply when creating a deployment with --push-to",
id="namespace_without_push_to",
),
],
)
def test_conflicting_flags_are_rejected(self, monkeypatch, flags, message):
@@ -890,3 +933,289 @@ class TestResolvePushedImageDigest:
frame_locals = captured["coro"].cr_frame.f_locals
assert "--config" not in frame_locals["args"]
captured["coro"].close()
class TestListener:
@pytest.mark.parametrize(
("resource", "expected"),
[
pytest.param(
{
"id": "listener-1",
"compute_id": "prod-cluster",
"compute_config": {"k8s_namespaces": ["agents", "agents-staging"]},
},
Listener("listener-1", "prod-cluster", ("agents", "agents-staging")),
id="reads_id_cluster_and_namespaces",
),
pytest.param(
{"id": "listener-1", "compute_id": "c", "compute_config": {}},
Listener("listener-1", "c", ()),
id="missing_namespaces",
),
pytest.param(
{"id": "listener-1", "compute_id": "c", "compute_config": None},
Listener("listener-1", "c", ()),
id="null_compute_config",
),
pytest.param(
{"id": "listener-1"},
Listener("listener-1", "", ()),
id="only_an_id",
),
],
)
def test_from_resource_reads_the_control_plane_shape(self, resource, expected):
assert Listener.from_resource(resource) == expected
ONE_NAMESPACE = Listener("listener-1", "prod-cluster", ("agents",))
TWO_NAMESPACES = Listener("listener-2", "multi-cluster", ("agents", "agents-staging"))
NO_NAMESPACE = Listener("listener-3", "broken-cluster", ())
class TestRequestedPlacement:
@pytest.mark.parametrize(
("request_", "listeners", "expected"),
[
pytest.param(
RequestedPlacement(), (), Unplaced(), id="no_listeners_no_request"
),
pytest.param(
RequestedPlacement(),
(ONE_NAMESPACE,),
OnListener("listener-1", "agents"),
id="uses_the_only_possible_answer",
),
pytest.param(
RequestedPlacement(k8s_namespace="agents-staging"),
(TWO_NAMESPACES,),
OnListener("listener-2", "agents-staging"),
id="namespace_alone_picks_the_only_listener",
),
],
)
def test_resolves_to_a_placement(self, request_, listeners, expected):
assert request_.among(listeners) == expected
@pytest.mark.parametrize(
("request_", "listeners", "message"),
[
pytest.param(
RequestedPlacement(listener_id="listener-1"),
(),
"no listeners",
id="workspace_has_no_listeners",
),
pytest.param(
RequestedPlacement(),
(ONE_NAMESPACE, TWO_NAMESPACES),
"--listener-id",
id="several_listeners_need_a_choice",
),
pytest.param(
RequestedPlacement(k8s_namespace="agents"),
(ONE_NAMESPACE, TWO_NAMESPACES),
"--listener-id",
id="namespace_alone_is_ambiguous_with_several_listeners",
),
pytest.param(
RequestedPlacement(k8s_namespace="agents"),
(),
"no listeners",
id="namespace_without_any_listener",
),
pytest.param(
RequestedPlacement(),
(TWO_NAMESPACES,),
"--k8s-namespace",
id="several_namespaces_need_a_choice",
),
],
)
def test_refuses_and_names_the_choices(self, request_, listeners, message):
with pytest.raises(click.UsageError, match=message):
request_.among(listeners)
def test_the_error_lists_every_listener_with_its_cluster_and_namespaces(self):
with pytest.raises(click.UsageError) as error:
RequestedPlacement().among((ONE_NAMESPACE, TWO_NAMESPACES))
assert "listener-1" in error.value.message
assert "prod-cluster" in error.value.message
assert "agents-staging" in error.value.message
@pytest.mark.parametrize(
("placement", "expected"),
[
pytest.param(Unplaced(), {}, id="unplaced_adds_nothing"),
pytest.param(
OnListener("listener-1", "agents"),
{
"listener_id": "listener-1",
"listener_config": {"k8s_namespace": "agents"},
},
id="placed_carries_listener_and_namespace",
),
],
)
def test_source_config_matches_the_control_plane_shape(self, placement, expected):
assert placement.source_config() == expected
def test_finding_a_deployment_by_name_narrows_the_search_for_every_server_version():
seen: dict = {}
def handler(req: httpx.Request) -> httpx.Response:
seen["params"] = dict(req.url.params)
return httpx.Response(
200,
json={"resources": [{"id": "dep-1", "name": "agent", "source": "github"}]},
)
client = HostBackendClient(
"https://api.example.com", "key", transport=httpx.MockTransport(handler)
)
found = find_deployment_by_name(client, "agent")
assert seen["params"] == {
"name": "agent",
"name_contains": "agent",
"limit": "100",
}
assert found == ExistingDeployment("dep-1", "github")
def test_a_server_that_ignores_the_exact_name_filter_never_matches_another_deployment():
client = HostBackendClient(
"https://api.example.com",
"key",
transport=httpx.MockTransport(
lambda req: httpx.Response(
200,
json={
"resources": [
{
"id": "dep-other",
"name": "another-teams-agent",
"source": "external_docker",
}
]
},
)
),
)
assert find_deployment_by_name(client, "brand-new-agent") is None
def test_a_full_page_without_a_match_refuses_to_claim_the_name_is_free():
page = [
{"id": f"dep-{index}", "name": f"other-agent-{index}"} for index in range(100)
]
client = HostBackendClient(
"https://api.example.com",
"key",
transport=httpx.MockTransport(
lambda req: httpx.Response(200, json={"resources": page})
),
)
with pytest.raises(click.ClickException, match="--deployment-id"):
find_deployment_by_name(client, "brand-new-agent")
def test_a_partial_page_without_a_match_means_the_name_is_free():
client = HostBackendClient(
"https://api.example.com",
"key",
transport=httpx.MockTransport(
lambda req: httpx.Response(
200, json={"resources": [{"id": "dep-1", "name": "other"}]}
)
),
)
assert find_deployment_by_name(client, "brand-new-agent") is None
@pytest.mark.parametrize(
"resource",
[
pytest.param({"compute_id": "c"}, id="no_id"),
pytest.param({"id": ""}, id="empty_id"),
],
)
def test_a_listener_without_an_id_is_refused(resource):
with pytest.raises(HostBackendError, match="without an id"):
Listener.from_resource(resource)
def test_a_deployment_id_with_listener_flags_is_refused_without_probing_docker(
monkeypatch,
):
def explode() -> tuple[bool, str | None]:
raise AssertionError("docker must not be probed for an argv-only conflict")
monkeypatch.setattr(deploy_mod, "can_build_locally", explode)
with pytest.raises(click.UsageError, match="--deployment-id"):
_select_source(
push_to="registry.example.com/app",
image=None,
image_name=None,
tag=None,
remote_build_flag=None,
placement=RequestedPlacement(listener_id="listener-1"),
selector=ById("dep-1"),
)
class TestPlacementOnAKnownListener:
@pytest.mark.parametrize(
("request_", "listener", "expected"),
[
pytest.param(
RequestedPlacement(listener_id="listener-1"),
ONE_NAMESPACE,
OnListener("listener-1", "agents"),
id="the_only_namespace_is_used",
),
pytest.param(
RequestedPlacement(listener_id="listener-2", k8s_namespace="agents"),
TWO_NAMESPACES,
OnListener("listener-2", "agents"),
id="the_chosen_namespace_is_used",
),
],
)
def test_places_on_the_listener(self, request_, listener, expected):
assert request_.on(listener) == expected
@pytest.mark.parametrize(
("request_", "listener", "message"),
[
pytest.param(
RequestedPlacement(listener_id="listener-2"),
TWO_NAMESPACES,
"--k8s-namespace",
id="several_namespaces_need_a_choice",
),
pytest.param(
RequestedPlacement(listener_id="listener-2", k8s_namespace="nope"),
TWO_NAMESPACES,
"does not serve namespace",
id="unknown_namespace",
),
pytest.param(
RequestedPlacement(listener_id="listener-3"),
NO_NAMESPACE,
"serves no namespaces",
id="listener_without_namespaces",
),
],
)
def test_refuses_and_names_the_namespaces(self, request_, listener, message):
with pytest.raises(click.UsageError, match=message):
request_.on(listener)
+148 -14
View File
@@ -79,19 +79,6 @@ def test_request_transport_error_raises():
c._request("GET", "/test")
def test_list_deployments_sends_query_params():
def handler(req: httpx.Request) -> httpx.Response:
assert req.url.path == "/v2/deployments"
assert req.url.params["name_contains"] == "my app"
return httpx.Response(200, json={"ok": True})
c = HostBackendClient(
"https://api.example.com", "test-key", transport=httpx.MockTransport(handler)
)
result = c.list_deployments("my app")
assert result == {"ok": True}
def _capturing_client(captured: dict) -> HostBackendClient:
def handler(req: httpx.Request) -> httpx.Response:
captured["body"] = req.read()
@@ -421,7 +408,7 @@ def test_injected_transport_receives_requests_under_the_prefixed_base_url():
transport=httpx.MockTransport(handler),
)
assert c.list_revisions("dep-1", limit=2) == {"ok": True}
assert c.list_revisions("dep-1", limit=2) == []
assert seen == {
"url": "https://smith.example.com/api-host/v2/deployments/dep-1/revisions?limit=2",
"api_key": "key",
@@ -540,9 +527,156 @@ CLOUD = ("https://api.host.langchain.com", "https://smith.langchain.com")
("https://eu.api.host.langchain.com", "https://eu.smith.langchain.com"),
id="regional_control_plane_maps_to_regional_dashboard",
),
pytest.param(
"https://dev.api.smith.langchain.com/",
None,
("https://dev.api.smith.langchain.com", "https://dev.smith.langchain.com"),
id="dev_smith_api_maps_to_dev_dashboard",
),
],
)
def test_control_plane_endpoints_resolve(host_url, langsmith_endpoint, expected):
endpoints = ControlPlaneEndpoints.resolve(host_url, langsmith_endpoint)
assert (endpoints.control_plane_url, endpoints.dashboard_url) == expected
@pytest.mark.parametrize(
("payload", "expected"),
[
pytest.param(
{"resources": [{"id": "a"}, {"id": "b"}]},
[{"id": "a"}, {"id": "b"}],
id="list_returns_the_resources",
),
pytest.param({"resources": []}, [], id="empty_list"),
pytest.param({}, [], id="missing_key"),
pytest.param({"resources": None}, [], id="null_resources"),
pytest.param(
{"resources": ["nope", {"id": "a"}]}, [{"id": "a"}], id="skips_non_objects"
),
pytest.param([], [], id="unexpected_envelope"),
],
)
def test_list_endpoints_return_resource_objects(payload, expected):
def handler(req: httpx.Request) -> httpx.Response:
return httpx.Response(200, json=payload)
c = HostBackendClient(
"https://api.example.com", "key", transport=httpx.MockTransport(handler)
)
assert c.list_deployments() == expected
def test_list_listeners_asks_for_a_full_page():
seen: dict = {}
def handler(req: httpx.Request) -> httpx.Response:
seen["url"] = str(req.url)
return httpx.Response(200, json={"resources": [{"id": "listener-1"}]})
c = HostBackendClient(
"https://api.example.com", "key", transport=httpx.MockTransport(handler)
)
assert c.list_listeners() == [{"id": "listener-1"}]
assert seen["url"] == "https://api.example.com/v2/listeners?limit=100"
@pytest.mark.parametrize(
("control_plane_url", "expected"),
[
pytest.param("https://api.host.langchain.com", True, id="cloud"),
pytest.param("https://eu.api.host.langchain.com", True, id="cloud_region"),
pytest.param("https://dev.api.host.langchain.com", True, id="cloud_dev"),
pytest.param("https://smith.example.com/api-host", False, id="self_hosted"),
pytest.param(
"https://corp.example.com/langsmith/api-host",
False,
id="self_hosted_prefix",
),
pytest.param("http://localhost:8080/api-host", False, id="local"),
pytest.param(
"https://evil-api.host.langchain.com", False, id="lookalike_needs_a_dot"
),
],
)
def test_is_cloud_recognises_the_managed_control_plane(control_plane_url, expected):
endpoints = ControlPlaneEndpoints.from_control_plane_url(control_plane_url)
assert endpoints.is_cloud is expected
@pytest.mark.parametrize(
("call", "expected_params"),
[
pytest.param(
lambda c: c.list_deployments(name="agent"),
{"name": "agent"},
id="exact_name_filters_server_side",
),
pytest.param(
lambda c: c.list_deployments(name_contains="age"),
{"name_contains": "age"},
id="substring_search_keeps_its_own_parameter",
),
pytest.param(
lambda c: c.list_deployments(),
{},
id="no_filter_sends_no_parameters",
),
pytest.param(
lambda c: c.list_deployments(
name="agent", name_contains="agent", limit=100
),
{"name": "agent", "name_contains": "agent", "limit": "100"},
id="both_filters_travel_together_for_older_servers",
),
],
)
def test_list_deployments_sends_one_name_filter(call, expected_params):
seen: dict = {}
def handler(req: httpx.Request) -> httpx.Response:
seen.update(dict(req.url.params))
return httpx.Response(200, json={"resources": []})
call(
HostBackendClient(
"https://api.example.com", "key", transport=httpx.MockTransport(handler)
)
)
assert seen == expected_params
@pytest.mark.parametrize(
("body", "expected"),
[
pytest.param(
{"detail": "Source configuration error: bad listener"},
"Source configuration error: bad listener",
id="fastapi_detail_is_unwrapped",
),
pytest.param(
{"detail": {"loc": ["body"], "msg": "nope"}},
None,
id="a_structured_detail_is_left_alone",
),
pytest.param({"other": "shape"}, None, id="an_unknown_shape_is_left_alone"),
],
)
def test_error_detail_is_readable(body, expected):
c = HostBackendClient(
"https://api.example.com",
"key",
transport=httpx.MockTransport(lambda req: httpx.Response(400, json=body)),
)
with pytest.raises(HostBackendError) as error:
c.get_deployment("dep-1")
assert error.value.detail == expected
if expected is not None:
assert error.value.message.endswith(expected)
+9 -55
View File
@@ -80,14 +80,12 @@ def get_updated_channels_from_tasks(
def get_delta_channels_from_all_channels(
channels: Mapping[str, BaseChannel],
*,
include_unavailable: bool = False,
) -> set[str]:
"""DeltaChannels to snapshot on the first update_state of a fresh thread or fork."""
"""DeltaChannels to snapshot on the first update_state of a fresh thread."""
return {
k
for k, ch in channels.items()
if isinstance(ch, DeltaChannel) and (include_unavailable or ch.is_available())
if isinstance(ch, DeltaChannel) and ch.is_available()
}
@@ -124,22 +122,15 @@ def create_checkpoint_plan_for_update_state_api(
parents: dict[str, Any],
saved_metadata: Mapping[str, Any] | None,
is_fresh_thread: bool,
is_fork: bool,
) -> tuple[set[str], dict[str, Any]]:
"""Return ``(channels_to_snapshot, metadata)`` for an update_state head.
A fork snapshots everything, like a fresh thread: its base also holds the
writes of the branch it abandons, so the ancestor walk must stop here.
"""
"""Return ``(channels_to_snapshot, metadata)`` for an update_state head."""
metadata: dict[str, Any] = {
"source": "update",
"step": step,
"parents": parents,
}
if is_fresh_thread or is_fork:
return get_delta_channels_from_all_channels(
channels, include_unavailable=is_fork
), metadata
if is_fresh_thread:
return get_delta_channels_from_all_channels(channels), metadata
new_counters = create_metadata_for_update_state_api(
channels,
@@ -155,34 +146,6 @@ def create_checkpoint_plan_for_update_state_api(
return channels_to_snapshot, metadata
def create_fork_checkpoint(
checkpoint: Checkpoint,
channels: Mapping[str, BaseChannel],
step: int,
*,
is_fork: bool,
get_next_version: GetNextVersion,
) -> Checkpoint:
"""``create_checkpoint`` for the update_state paths that skip the plan.
The fork has to be sealed by its first checkpoint: any later superstep
has already rebuilt its delta channels through the shared base. These
paths never write the delta channel, so its version must be bumped here
or ``put`` drops the blob; derive ``new_versions`` from the result.
"""
if not is_fork:
return create_checkpoint(checkpoint, channels, step)
return create_checkpoint(
checkpoint,
channels,
step,
get_next_version=get_next_version,
channels_to_snapshot=get_delta_channels_from_all_channels(
channels, include_unavailable=True
),
)
def create_checkpoint(
checkpoint: Checkpoint,
channels: Mapping[str, BaseChannel] | None,
@@ -211,23 +174,14 @@ def create_checkpoint(
values = {}
channel_versions = dict(checkpoint["channel_versions"])
for k in channels:
ch = channels[k]
if k not in channel_versions:
# A forced snapshot of a never-written channel still has to
# land to stop the ancestor walk, and `put` only stores blobs
# for versioned channels.
if k in channels_to_snapshot and get_next_version is not None:
channel_versions[k] = get_next_version(None, None)
values[k] = _DeltaSnapshot(
ch.get() if ch.is_available() else ch.typ()
)
continue
ch = channels[k]
if k in channels_to_snapshot:
# Callers force a full snapshot blob here: exit mode when a
# delta channel reaches its snapshot cadence, update_state on
# a fresh thread (no ancestor to replay writes from), and a
# fork. The manual version-bump below only applies to the
# exit-mode case.
# delta channel reaches its snapshot cadence, and update_state
# on a fresh thread (no ancestor to replay writes from). The
# manual version-bump below only applies to the exit-mode case.
#
# In exit mode, the snapshot decision is deferred to exit
# time (intermediate steps have do_checkpoint=False). The
+11 -22
View File
@@ -222,16 +222,10 @@ class PregelLoop:
# under the saver's `ORDER BY task_id, idx` sorting.
_exit_delta_writes: list[tuple[int, str, str, Any]] | None = None
# Delta channels that must snapshot at the next checkpoint, whatever their
# cadence counters say:
# * an Overwrite arrived since the last checkpoint, so sparse replay has to
# start from the post-overwrite value;
# * this run forked off an explicitly addressed checkpoint. That base also
# holds the writes of the branch the fork abandons, and nothing records
# which child consumed which, so the ancestor walk must stop inside the
# fork. Any addressed checkpoint counts, because telling a real fork
# apart would mean trusting the base's `pending_writes` to be complete.
_delta_channels_forced_snapshot: set[str]
# Delta channels that saw an Overwrite since the last checkpoint. These
# channels must snapshot after live update applies overwrite semantics so
# sparse replay starts from the same post-overwrite value.
_delta_channels_with_overwrite: set[str]
# The checkpoint_config that points at the parent loaded at `__enter__`
# (or the synthetic-empty checkpoint, on first run). We capture it
@@ -375,13 +369,6 @@ class PregelLoop:
if self.config[CONF].get(CONFIG_KEY_CHECKPOINT_NS)
else ()
)
# Value, not key presence like `is_replaying`: subgraph task configs
# always carry an explicit `None` checkpoint_id.
self._delta_channels_forced_snapshot = (
{k for k, spec in specs.items() if isinstance(spec, DeltaChannel)}
if self.checkpoint_config[CONF].get(CONFIG_KEY_CHECKPOINT_ID)
else set()
)
self.prev_checkpoint_config = None
runtime = self.config[CONF].get(CONFIG_KEY_RUNTIME)
self.control = runtime.control if isinstance(runtime, Runtime) else None
@@ -696,7 +683,7 @@ class PregelLoop:
def after_tick(self) -> None:
# finish superstep
writes = [w for t in self.tasks.values() for w in t.writes]
self._delta_channels_forced_snapshot.update(
self._delta_channels_with_overwrite.update(
ch
for ch, v in writes
if isinstance(self.specs.get(ch), DeltaChannel) and _get_overwrite(v)[0]
@@ -1004,7 +991,7 @@ class PregelLoop:
manager=None,
updated_channels=updated_channels,
)
self._delta_channels_forced_snapshot.update(
self._delta_channels_with_overwrite.update(
c
for c, v in input_writes
if isinstance(self.specs.get(c), DeltaChannel) and _get_overwrite(v)[0]
@@ -1149,7 +1136,7 @@ class PregelLoop:
# create new checkpoint
channels_to_snapshot = (
delta_channels_to_snapshot(self.channels, new_counters)
| self._delta_channels_forced_snapshot
| self._delta_channels_with_overwrite
if do_checkpoint
else set()
)
@@ -1167,7 +1154,7 @@ class PregelLoop:
for k in channels_to_snapshot:
new_counters[k] = (0, 0)
if do_checkpoint:
self._delta_channels_forced_snapshot.difference_update(channels_to_snapshot)
self._delta_channels_with_overwrite.difference_update(channels_to_snapshot)
non_zero = {k: v for k, v in new_counters.items() if v != (0, 0)}
if non_zero:
self.checkpoint_metadata["counters_since_delta_snapshot"] = non_zero
@@ -1252,7 +1239,7 @@ class PregelLoop:
)
channels_to_snapshot = (
delta_channels_to_snapshot(self.channels, counters)
| self._delta_channels_forced_snapshot
| self._delta_channels_with_overwrite
)
pending = [
@@ -1697,6 +1684,7 @@ class SyncPregelLoop(PregelLoop, AbstractContextManager):
)
self._delta_write_futs = []
self._error_handler_write_futs = []
self._delta_channels_with_overwrite = set()
self._exit_delta_writes = (
[] if self.durability == "exit" and self.checkpointer is not None else None
)
@@ -1954,6 +1942,7 @@ class AsyncPregelLoop(PregelLoop, AbstractAsyncContextManager):
)
self._delta_write_futs = []
self._error_handler_write_futs = []
self._delta_channels_with_overwrite = set()
self._exit_delta_writes = (
[] if self.durability == "exit" and self.checkpointer is not None else None
)
+12 -85
View File
@@ -108,7 +108,6 @@ from langgraph.callbacks import (
get_sync_graph_callback_manager_for_config,
)
from langgraph.channels.base import BaseChannel
from langgraph.channels.delta import DeltaChannel
from langgraph.channels.topic import Topic
from langgraph.config import get_config
from langgraph.constants import END
@@ -134,7 +133,6 @@ from langgraph.pregel._checkpoint import (
copy_checkpoint,
create_checkpoint,
create_checkpoint_plan_for_update_state_api,
create_fork_checkpoint,
empty_checkpoint,
get_updated_channels_from_tasks,
)
@@ -1639,21 +1637,8 @@ class Pregel(
else:
raise ValueError(f"Subgraph {recast} not found")
# Read once from the caller's config: every later superstep receives
# the config of the checkpoint just written, which always names one.
# Cleared by the first checkpoint that carries the snapshots, which
# `__copy__` does not write.
fork_pending: set[str] = (
{k for k, v in self.channels.items() if isinstance(v, DeltaChannel)}
if config[CONF].get(CONFIG_KEY_CHECKPOINT_ID)
else set()
)
def perform_superstep(
input_config: RunnableConfig,
updates: Sequence[StateUpdate],
*,
is_fork: bool,
input_config: RunnableConfig, updates: Sequence[StateUpdate]
) -> RunnableConfig:
# get last checkpoint
config = ensure_config(self.config, input_config)
@@ -1741,17 +1726,9 @@ class Pregel(
self.trigger_to_nodes,
)
# save checkpoint
next_checkpoint = create_fork_checkpoint(
checkpoint,
channels,
step,
is_fork=is_fork,
get_next_version=checkpointer.get_next_version,
)
fork_pending.difference_update(next_checkpoint["channel_values"])
next_config = checkpointer.put(
checkpoint_config,
next_checkpoint,
create_checkpoint(checkpoint, channels, step),
{
"source": "update",
"step": step + 1,
@@ -1759,7 +1736,7 @@ class Pregel(
},
get_new_channel_versions(
checkpoint_previous_versions,
next_checkpoint["channel_versions"],
checkpoint["channel_versions"],
),
)
return patch_checkpoint_map(
@@ -1788,17 +1765,9 @@ class Pregel(
if saved and saved.metadata.get("step") is not None
else -1
)
next_checkpoint = create_fork_checkpoint(
checkpoint,
channels,
next_step,
is_fork=is_fork,
get_next_version=checkpointer.get_next_version,
)
fork_pending.difference_update(next_checkpoint["channel_values"])
next_config = checkpointer.put(
checkpoint_config,
next_checkpoint,
create_checkpoint(checkpoint, channels, next_step),
{
"source": "input",
"step": next_step,
@@ -1808,7 +1777,7 @@ class Pregel(
},
get_new_channel_versions(
checkpoint_previous_versions,
next_checkpoint["channel_versions"],
checkpoint["channel_versions"],
),
)
@@ -1904,7 +1873,6 @@ class Pregel(
return perform_superstep(
patch_checkpoint_map(next_config, saved.metadata),
[item for lst in user_group_by.values() for item in lst],
is_fork=is_fork,
)
return patch_checkpoint_map(next_config, saved.metadata)
@@ -2052,7 +2020,6 @@ class Pregel(
parents=saved.metadata.get("parents", {}) if saved else {},
saved_metadata=saved.metadata if saved else None,
is_fresh_thread=saved is None,
is_fork=is_fork,
)
)
checkpoint = create_checkpoint(
@@ -2065,8 +2032,6 @@ class Pregel(
else None,
channels_to_snapshot=channels_to_snapshot,
)
if is_fork:
fork_pending.difference_update(checkpoint["channel_values"])
next_config = checkpointer.put(
checkpoint_config,
checkpoint,
@@ -2085,9 +2050,7 @@ class Pregel(
config, {CONFIG_KEY_THREAD_ID: str(config[CONF][CONFIG_KEY_THREAD_ID])}
)
for superstep in supersteps:
current_config = perform_superstep(
current_config, superstep, is_fork=bool(fork_pending)
)
current_config = perform_superstep(current_config, superstep)
return current_config
async def abulk_update_state(
@@ -2140,21 +2103,8 @@ class Pregel(
else:
raise ValueError(f"Subgraph {recast} not found")
# Read once from the caller's config: every later superstep receives
# the config of the checkpoint just written, which always names one.
# Cleared by the first checkpoint that carries the snapshots, which
# `__copy__` does not write.
fork_pending: set[str] = (
{k for k, v in self.channels.items() if isinstance(v, DeltaChannel)}
if config[CONF].get(CONFIG_KEY_CHECKPOINT_ID)
else set()
)
async def aperform_superstep(
input_config: RunnableConfig,
updates: Sequence[StateUpdate],
*,
is_fork: bool,
input_config: RunnableConfig, updates: Sequence[StateUpdate]
) -> RunnableConfig:
# get last checkpoint
config = ensure_config(self.config, input_config)
@@ -2240,25 +2190,16 @@ class Pregel(
self.trigger_to_nodes,
)
# save checkpoint
next_checkpoint = create_fork_checkpoint(
checkpoint,
channels,
step,
is_fork=is_fork,
get_next_version=checkpointer.get_next_version,
)
fork_pending.difference_update(next_checkpoint["channel_values"])
next_config = await checkpointer.aput(
checkpoint_config,
next_checkpoint,
create_checkpoint(checkpoint, channels, step),
{
"source": "update",
"step": step + 1,
"parents": saved.metadata.get("parents", {}) if saved else {},
},
get_new_channel_versions(
checkpoint_previous_versions,
next_checkpoint["channel_versions"],
checkpoint_previous_versions, checkpoint["channel_versions"]
),
)
return patch_checkpoint_map(
@@ -2287,17 +2228,9 @@ class Pregel(
if saved and saved.metadata.get("step") is not None
else -1
)
next_checkpoint = create_fork_checkpoint(
checkpoint,
channels,
next_step,
is_fork=is_fork,
get_next_version=checkpointer.get_next_version,
)
fork_pending.difference_update(next_checkpoint["channel_values"])
next_config = await checkpointer.aput(
checkpoint_config,
next_checkpoint,
create_checkpoint(checkpoint, channels, next_step),
{
"source": "input",
"step": next_step,
@@ -2307,7 +2240,7 @@ class Pregel(
},
get_new_channel_versions(
checkpoint_previous_versions,
next_checkpoint["channel_versions"],
checkpoint["channel_versions"],
),
)
@@ -2402,7 +2335,6 @@ class Pregel(
return await aperform_superstep(
patch_checkpoint_map(next_config, saved.metadata),
[item for lst in user_group_by.values() for item in lst],
is_fork=is_fork,
)
return patch_checkpoint_map(
@@ -2548,7 +2480,6 @@ class Pregel(
parents=saved.metadata.get("parents", {}) if saved else {},
saved_metadata=saved.metadata if saved else None,
is_fresh_thread=saved is None,
is_fork=is_fork,
)
)
checkpoint = create_checkpoint(
@@ -2561,8 +2492,6 @@ class Pregel(
else None,
channels_to_snapshot=channels_to_snapshot,
)
if is_fork:
fork_pending.difference_update(checkpoint["channel_values"])
next_config = await checkpointer.aput(
checkpoint_config,
checkpoint,
@@ -2580,9 +2509,7 @@ class Pregel(
config, {CONFIG_KEY_THREAD_ID: str(config[CONF][CONFIG_KEY_THREAD_ID])}
)
for superstep in supersteps:
current_config = await aperform_superstep(
current_config, superstep, is_fork=bool(fork_pending)
)
current_config = await aperform_superstep(current_config, superstep)
return current_config
def update_state(
+3 -5
View File
@@ -85,13 +85,11 @@ class MemorySaverAssertImmutable(InMemorySaver):
)
== saved
), config["configurable"]["checkpoint_ns"]
next_config = super().put(config, checkpoint, metadata, new_versions)
# Read back, not the object handed in: a DeltaChannel a step did not
# write is refilled on read from the blob its inherited version points at.
self.storage_for_copies[thread_id][checkpoint_ns][checkpoint["id"]] = (
self.serde.dumps_typed(super().get(next_config))
self.serde.dumps_typed(checkpoint)
)
return next_config
# call super to write checkpoint
return super().put(config, checkpoint, metadata, new_versions)
class MemorySaverNoPending(InMemorySaver):
@@ -1,346 +0,0 @@
"""Forking a thread must not replay the abandoned branch into the fork.
Every graph carries a ``DeltaChannel`` and a plain reducer channel fed the same
values; the plain channel needs no replay, so it is the oracle.
"""
from collections.abc import Sequence
from operator import add
from typing import Annotated, Any
import pytest
from langchain_core.runnables import RunnableConfig
from langgraph.checkpoint.base import BaseCheckpointSaver
from langgraph.checkpoint.serde.types import _DeltaSnapshot
from typing_extensions import TypedDict
from langgraph._internal._constants import INPUT
from langgraph.channels.delta import DeltaChannel
from langgraph.graph import END, StateGraph
from langgraph.types import Durability, StateSnapshot, StateUpdate
pytestmark = pytest.mark.anyio
def _append(current: list | None, writes: Sequence[Any]) -> list:
out = list(current or [])
for write in writes:
out.extend(write if isinstance(write, list) else [write])
return out
class _State(TypedDict):
log: Annotated[list, DeltaChannel(_append, snapshot_frequency=1000)]
plain: Annotated[list, add]
other: Annotated[list, add]
def _build(checkpointer: BaseCheckpointSaver, tag: str) -> Any:
def node(state: _State) -> dict:
return {"log": [f"{tag}-out"], "plain": [f"{tag}-out"]}
builder = StateGraph(_State)
builder.add_node("n", node)
builder.set_entry_point("n")
builder.set_finish_point("n")
return builder.compile(checkpointer=checkpointer)
def _build_without_delta_writes(checkpointer: BaseCheckpointSaver, tag: str) -> Any:
def node(state: _State) -> dict:
return {"other": [f"{tag}-other"]}
builder = StateGraph(_State)
builder.add_node("n", node)
builder.set_entry_point("n")
builder.set_finish_point("n")
return builder.compile(checkpointer=checkpointer)
def _thread(thread_id: str) -> RunnableConfig:
return {"configurable": {"thread_id": thread_id}}
def _at(config: RunnableConfig, snapshot: StateSnapshot) -> RunnableConfig:
return {
"configurable": {
**config["configurable"],
"checkpoint_ns": "",
"checkpoint_id": snapshot.config["configurable"]["checkpoint_id"],
}
}
def _input(marker: str) -> dict:
return {"log": [marker], "plain": [marker]}
def _snapshotted_checkpoints(
checkpointer: BaseCheckpointSaver, config: RunnableConfig
) -> list[str]:
return [
tuple_.config["configurable"]["checkpoint_id"]
for tuple_ in checkpointer.list(config)
if isinstance(tuple_.checkpoint["channel_values"].get("log"), _DeltaSnapshot)
]
def _assert_fork_is_clean(state: StateSnapshot, abandoned: str) -> None:
assert state.values["log"] == state.values["plain"], (
f"delta channel diverged from the plain channel: "
f"{state.values['log']} != {state.values['plain']}"
)
assert abandoned not in state.values["log"], (
f"{abandoned!r} belongs to the branch the fork replaced, "
f"but was replayed into {state.values['log']}"
)
def test_fork_by_invoke(
sync_checkpointer: BaseCheckpointSaver, durability: Durability
) -> None:
config = _thread("t")
_build(sync_checkpointer, "first").invoke(
_input("in-1"), config, durability=durability
)
graph = _build(sync_checkpointer, "second")
graph.invoke(_input("in-2"), config, durability=durability)
abandoned_head = graph.get_state(config)
base = next(
snapshot
for snapshot in graph.get_state_history(config)
if "in-2" not in snapshot.values["log"]
)
_build(sync_checkpointer, "third").invoke(
_input("in-3"), _at(config, base), durability=durability
)
state = graph.get_state(config)
_assert_fork_is_clean(state, "in-2")
assert state.values["log"] == [*base.values["log"], "in-3", "third-out"]
abandoned = graph.get_state(abandoned_head.config).values
assert abandoned["log"] == abandoned["plain"] == abandoned_head.values["log"]
async def test_afork_by_invoke(
async_checkpointer: BaseCheckpointSaver, durability: Durability
) -> None:
config = _thread("t")
await _build(async_checkpointer, "first").ainvoke(
_input("in-1"), config, durability=durability
)
graph = _build(async_checkpointer, "second")
await graph.ainvoke(_input("in-2"), config, durability=durability)
abandoned_head = await graph.aget_state(config)
base = await anext(
snapshot
async for snapshot in graph.aget_state_history(config)
if "in-2" not in snapshot.values["log"]
)
await _build(async_checkpointer, "third").ainvoke(
_input("in-3"), _at(config, base), durability=durability
)
state = await graph.aget_state(config)
_assert_fork_is_clean(state, "in-2")
assert state.values["log"] == [*base.values["log"], "in-3", "third-out"]
abandoned = (await graph.aget_state(abandoned_head.config)).values
assert abandoned["log"] == abandoned["plain"] == abandoned_head.values["log"]
def test_fork_off_checkpoint_before_first_input(
sync_checkpointer: BaseCheckpointSaver, durability: Durability
) -> None:
config = _thread("t")
graph = _build(sync_checkpointer, "first")
graph.invoke(_input("in-1"), config, durability=durability)
root = list(graph.get_state_history(config))[-1]
assert root.values["log"] == []
_build(sync_checkpointer, "third").invoke(
_input("in-9"), _at(config, root), durability=durability
)
state = graph.get_state(config)
_assert_fork_is_clean(state, "in-1")
assert state.values["log"] == ["in-9", "third-out"]
async def test_afork_off_checkpoint_before_first_input(
async_checkpointer: BaseCheckpointSaver, durability: Durability
) -> None:
config = _thread("t")
graph = _build(async_checkpointer, "first")
await graph.ainvoke(_input("in-1"), config, durability=durability)
root = [snapshot async for snapshot in graph.aget_state_history(config)][-1]
assert root.values["log"] == []
await _build(async_checkpointer, "third").ainvoke(
_input("in-9"), _at(config, root), durability=durability
)
state = await graph.aget_state(config)
_assert_fork_is_clean(state, "in-1")
assert state.values["log"] == ["in-9", "third-out"]
def test_fork_by_update_state(sync_checkpointer: BaseCheckpointSaver) -> None:
config = _thread("t")
_build(sync_checkpointer, "first").invoke(_input("in-1"), config)
graph = _build(sync_checkpointer, "second")
graph.invoke(_input("in-2"), config)
base = next(
snapshot
for snapshot in graph.get_state_history(config)
if "in-2" not in snapshot.values["log"]
)
forked = graph.update_state(_at(config, base), _input("patched"))
state = graph.get_state(forked)
_assert_fork_is_clean(state, "in-2")
assert state.values["log"] == [*base.values["log"], "patched"]
async def test_afork_by_update_state(
async_checkpointer: BaseCheckpointSaver,
) -> None:
config = _thread("t")
await _build(async_checkpointer, "first").ainvoke(_input("in-1"), config)
graph = _build(async_checkpointer, "second")
await graph.ainvoke(_input("in-2"), config)
base = await anext(
snapshot
async for snapshot in graph.aget_state_history(config)
if "in-2" not in snapshot.values["log"]
)
forked = await graph.aupdate_state(_at(config, base), _input("patched"))
state = await graph.aget_state(forked)
_assert_fork_is_clean(state, "in-2")
assert state.values["log"] == [*base.values["log"], "patched"]
def test_unaddressed_run_keeps_snapshot_cadence(
sync_checkpointer: BaseCheckpointSaver, durability: Durability
) -> None:
config = _thread("t")
graph = _build(sync_checkpointer, "first")
graph.invoke(_input("in-1"), config, durability=durability)
graph.invoke(_input("in-2"), config, durability=durability)
assert not _snapshotted_checkpoints(sync_checkpointer, config)
def test_fork_before_first_value_when_fork_never_writes_the_channel(
sync_checkpointer: BaseCheckpointSaver, durability: Durability
) -> None:
config = _thread("t")
graph = _build(sync_checkpointer, "first")
graph.invoke(_input("in-1"), config, durability=durability)
root = list(graph.get_state_history(config))[-1]
assert root.values["log"] == []
_build_without_delta_writes(sync_checkpointer, "third").invoke(
{"other": ["in-9"]}, _at(config, root), durability=durability
)
state = graph.get_state(config)
_assert_fork_is_clean(state, "in-1")
assert state.values["log"] == []
async def test_afork_before_first_value_when_fork_never_writes_the_channel(
async_checkpointer: BaseCheckpointSaver, durability: Durability
) -> None:
config = _thread("t")
graph = _build(async_checkpointer, "first")
await graph.ainvoke(_input("in-1"), config, durability=durability)
root = [snapshot async for snapshot in graph.aget_state_history(config)][-1]
assert root.values["log"] == []
await _build_without_delta_writes(async_checkpointer, "third").ainvoke(
{"other": ["in-9"]}, _at(config, root), durability=durability
)
state = await graph.aget_state(config)
_assert_fork_is_clean(state, "in-1")
assert state.values["log"] == []
def test_fork_before_first_value_by_bulk_update(
sync_checkpointer: BaseCheckpointSaver,
) -> None:
config = _thread("t")
graph = _build(sync_checkpointer, "first")
graph.invoke(_input("in-1"), config)
root = list(graph.get_state_history(config))[-1]
assert root.values["log"] == []
forked = graph.bulk_update_state(
_at(config, root),
[
[StateUpdate({"other": ["s1"]}, "n")],
[StateUpdate(_input("s2"), "n")],
],
)
state = graph.get_state(forked)
_assert_fork_is_clean(state, "in-1")
assert state.values["log"] == ["s2"]
@pytest.mark.parametrize("first_as_node", [INPUT, END, "__copy__"])
def test_fork_by_bulk_update_whose_first_superstep_skips_the_plan(
sync_checkpointer: BaseCheckpointSaver, first_as_node: str
) -> None:
config = _thread("t")
_build(sync_checkpointer, "first").invoke(_input("in-1"), config)
graph = _build(sync_checkpointer, "second")
graph.invoke(_input("in-2"), config)
base = next(
snapshot
for snapshot in graph.get_state_history(config)
if "in-2" not in snapshot.values["log"]
)
first = (
StateUpdate(_input("first-step"), first_as_node)
if first_as_node == INPUT
else StateUpdate(None, first_as_node)
)
forked = graph.bulk_update_state(
_at(config, base),
[[first], [StateUpdate(_input("second-step"), "n")]],
)
state = graph.get_state(forked)
assert state.values["log"] == state.values["plain"], (
f"delta channel diverged from the plain channel: "
f"{state.values['log']} != {state.values['plain']}"
)
def test_unaddressed_bulk_update_keeps_snapshot_cadence(
sync_checkpointer: BaseCheckpointSaver,
) -> None:
config = _thread("t")
graph = _build(sync_checkpointer, "first")
graph.invoke(_input("in-1"), config)
graph.bulk_update_state(
config,
[[StateUpdate(_input(f"u{i}"), "n")] for i in range(4)],
)
assert not _snapshotted_checkpoints(sync_checkpointer, config)