What happened:
UploadChunk creates multiple writes to the ChunkRef datastore entity. This is because a ChunkRef is created initially and then eventually marked as Ready after the blob is uploaded to GCS.
At best this is inefficient, and at worst causes contention.
What you expected to happen:
Only one write is made to the chunkref entity.
How to reproduce it (as minimally and precisely as possible):
Upload a Chunk
Anything else we need to know?:
We can remove this by creating the ChunkRef after the blob is created. If this upload fails, we can cancel the blob write.
What happened: UploadChunk creates multiple writes to the ChunkRef datastore entity. This is because a ChunkRef is created initially and then eventually marked as Ready after the blob is uploaded to GCS.
At best this is inefficient, and at worst causes contention.
What you expected to happen: Only one write is made to the chunkref entity.
How to reproduce it (as minimally and precisely as possible): Upload a Chunk
Anything else we need to know?:
We can remove this by creating the ChunkRef after the blob is created. If this upload fails, we can cancel the blob write.
Environment:
kubectl version):