googleapis / googleapis/python-aiplatform

gcp vertex ai engine concurrent import requests limit makes rag.import_files_async call almost pointless if uploading to multiple corpora

Abierto
#5,424 1 comentario 0 reacciones 0 asignados Ver en GitHub
api: vertex-ai
Lenguaje dominante
Python
Estrellas
905
Forks
465
Merge medio
1 d 13 h
PR fusionados (30 d)
44

Descripción

The code below works to get the first 50 files from `bucket_dir_1` and then does not get anymore, no does it get any other files at all from any of the other `bucket_dirs`. the vertex ai rag engine says that up to 10K files can be imported at one time so I've no idea why I'd be hitting a limit.

```
corpora_ids_and_display_names = [('bucket_dir_1', 'projects/$PROJECT_NUMBER/locations/us-central1/ragCorpora/$CORPUS_ID_1'), ('bucket_dir_2', 'projects/$PROJECT_NUMBER/locations/us-central1/ragCorpora/$CORPUS_ID_2')]

async def import_files_async(corpus_id_and_display_name: tuple):
try:
response = await rag.import_files_async(
corpus_name=corpus_id_and_display_name[1],
paths=[f"gs://some-bucket/{corpus_id_and_display_name[0]}"],
)
return response
except Exception as e:
print(f"error: {e}")

async def create_import_files_task_group_async():
coros = [import_files_async(corpus_id_and_display_name) for corpus_id_and_display_name in corpora_ids_and_display_names]
import_files_task_group = [asyncio.create_task(coro) for coro in coros]
results = await asyncio.gather(*import_files_task_group, return_exceptions=True)
print(f"results: {results}")

asyncio.run(create_import_files_task_group_async())
```

```
# final asyncio.gather/asyncio.run response -- no errors
results: [, , , , ]
```

```
# works to list corpora hence corpora paths are correct
for corpus_id_and_display_name in corpora_ids_and_display_names:
print(rag.list_files(corpus_name=corpus_id_and_display_name[1]))
```

```
# works to list all files in each bucket path hence import_files_async
# has valid paths argument
for corpus_id_and_display_name in corpora_ids_and_display_names:
subprocess.run(['gcloud storage ls', f"gs://bucket/{corpus_id_and_display_name[0]}"])
```

Guía de contribución

Abrir la guía de contribución

Evaluación

Este issue todavía no se ha evaluado.

Recibe los nuevos issues en tu correo

Un resumen breve de issues de GitHub para principiantes.