GCP的Vertex AI（AI平台）PipeLineserviceClient给出了不完美的错误

发布于 2025-01-29 05:46:02 字数 577 浏览 6 评论 0 原文

尝试使用 pipelineserviceclient list_pipeline_jobs 方法给定here, I get the following error:

_InactiveRpcError: <_InactiveRpcError of RPC that terminated with:
status = StatusCode.UNIMPLEMENTED
details = "Received http2 header with status: 404"
...

How is the API unimplemented, how do I resolve this?

原文

When trying to list pipelines with PipelineServiceClient list_pipeline_jobs method as given here, I get the following error:

_InactiveRpcError: <_InactiveRpcError of RPC that terminated with:
status = StatusCode.UNIMPLEMENTED
details = "Received http2 header with status: 404"
...

How is the API unimplemented, how do I resolve this?

分享到QQ

分享到微博

如果你对这篇内容有疑问，欢迎到本站社区发帖提问参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。

发布评论

需要登录才能够评论，你可以免费注册一个本站的账号。

日久见人心 2025-02-05 05:46:02

我们这里缺少两件事：

我们需要使用具有正确端点的客户端选项初始化Pipelineserviceclient。端点列表在这里。另外，它应该是端点的格式：端口
我们需要将正确的请求提供给 list_pipeline_jobs 。

示例：对于 Asia-South1 区域，我们将写一个这样的呼叫：

import google.cloud.aiplatform as aip
from google.api_core.client_options import ClientOptions

aip.init(project='<PROJECT_NAME>',
         staging_bucket='PROJECT_BUCKET',
         location='asia-south1')

options = ClientOptions(
    api_endpoint="asia-south1-aiplatform.googleapis.com:443")

cli = aip.gapic.PipelineServiceClient(client_options=options)

parent = cli.common_location_path(project='<PROJECT_NAME>', location='asia-south1')
request = aip.gapic.ListPipelineJobsRequest(
    request=request
)

pipeline_jobs_pager = cli.list_pipeline_jobs(parent=parent) # returns a pager object

for page in pipeline_jobs_pager.pages:
  for pipeline_job in page.pipeline_jobs:
    print(pipeline_job.name)

注意： gapic 可以用 v1 或 v1beta1 调用。

There are two things we are missing here:

We need to initialize PipelineServiceClient with client options with the correct endpoint. List of endpoints are given here. Also it should be in the format of endpoint:port
We need to give the correct request to the list_pipeline_jobs.

Example: for asia-south1 region, we would write a call like this:

import google.cloud.aiplatform as aip
from google.api_core.client_options import ClientOptions

aip.init(project='<PROJECT_NAME>',
         staging_bucket='PROJECT_BUCKET',
         location='asia-south1')

options = ClientOptions(
    api_endpoint="asia-south1-aiplatform.googleapis.com:443")

cli = aip.gapic.PipelineServiceClient(client_options=options)

parent = cli.common_location_path(project='<PROJECT_NAME>', location='asia-south1')
request = aip.gapic.ListPipelineJobsRequest(
    request=request
)

pipeline_jobs_pager = cli.list_pipeline_jobs(parent=parent) # returns a pager object

for page in pipeline_jobs_pager.pages:
  for pipeline_job in page.pipeline_jobs:
    print(pipeline_job.name)

Note: gapic can be replaced with v1 or v1beta1 calls.