GCP Vertex AI RAG Corpus创建失败及初始化异常技术咨询

问题1：携带Vertex Vector Search信息创建ragCorpus失败 #

• 请求负载：

{"display_name":"Rag corpus","description":"Rag Corpus to store in vector search","vectorDbConfig":{"vertexVectorSearch":
{"index":"projects/<projectid>/locations/<location>/indexes/<indexid>",
"index_endpoint":"projects/<projectid>/locations/<location>/indexEndpoints/<indexEndpoints>"},"ragEmbeddingModelConfig":
{"vertexPredictionEndpoint":{"endpoint":"projects/<projectid>/locations/<location>/publishers/google/models/text-embedding-005"}}}}

• 预期行为：ragCorpus应创建成功
• 实际行为：请求失败，无法创建语料库

问题2：空vertexVectorSearch字段创建ragCorpus成功，但始终处于INITIALIZATION状态 #

• 请求负载：

{"display_name":"Rag corpus","description":"Rag Corpus to store in vector search",
"vectorDbConfig":{"vertexVectorSearch":{},"ragEmbeddingModelConfig":{"vertexPredictionEndpoint":
{"endpoint":"projects/<projectid>/locations/<location>/publishers/google/models/text-embedding-005"}}}}

• 预期行为：ragCorpus应完成初始化并转为ACTIVE状态
• 实际行为：ragCorpus已创建但始终处于INITIALIZATION状态，可能存在后端服务或配置缺失问题

问题3：更换为text-embedding-003模型仍无法创建ragCorpus #

内容的提问来源于stack exchange，提问作者suraj soni