Vertex AI Agent Engine Update Error (Aborted 409)

[ktk94]

from vertexai import agent_engines
import vertexai

vertexai.init(project = "my-project",location="asia-northeast3",staging_bucket="gs://asdasd"

from google.adk.agents import LlmAgent
x = LlmAgent(
 name = 'test',model = 'gemini-2.5-flash',instruction = 'hello'
)

agent_engines.update(
 name = "projects/125125/locations/asia-northeast3/reasoningEngines/engine_id"
 agent_engins = x
)

Error : Aborted 409
The error occurred while deploying an empty agent to the agent engine and updating it.
It is confirmed that 3 files are created/updated to Bucket, but the agent is not updated.

[adk-bot]

Response from ADK Answering Agent (experimental, answer may be inaccurate)

TLDR: The agent_engines.update() method expects an agent_engines.AdkApp object, not a raw LlmAgent. You need to wrap your agent in an AdkApp before deploying. Also, the parameter name is agent_engine, not agent_engins.

---

Hello! The "Aborted 409" error you're seeing is likely because the agent isn't being packaged correctly for deployment to the Vertex AI Agent Engine. [1]

When deploying or updating an agent, you must first wrap your ADK agent object (in your case, the LlmAgent) inside an agent_engines.AdkApp instance. This AdkApp object is then passed to the update method. [1]

Here is a revised version of your code that demonstrates the correct approach:

import vertexai
from vertexai import agent_engines
from google.adk.agents import LlmAgent

# Initialize Vertex AI
vertexai.init(
    project="my-project",
    location="asia-northeast3",
    staging_bucket="gs://asdasd"
)

# 1. Define your ADK agent
my_adk_agent = LlmAgent(
    name='test',
    model='gemini-2.5-flash',
    instruction='hello'
)

# 2. Wrap the agent in an AdkApp object for deployment
app_to_deploy = agent_engines.AdkApp(
    agent=my_adk_agent
)

# 3. Call update with the AdkApp object and correct parameter name
agent_engines.update(
    name="projects/125125/locations/asia-northeast3/reasoningEngines/engine_id",
    agent_engine=app_to_deploy  # Correct parameter name is 'agent_engine'
)

The 409 Conflict error often happens when the request payload is not in the expected format, causing a conflict with the state of the target resource on the server. By using AdkApp, you are packaging your agent in the format the Agent Engine service requires for updates. [1]

Please try this updated structure. It should resolve the deployment error.

[1] https://google.github.io/adk-docs/deploy/agent-engine/