Open tshfc opened 1 year ago
Ouch. This is obviously painful, sorry about that.
In general it is NOT safe to pass errors messages from the block adaptor through to the external user: they might be (very) misleading, and possibly may even reveal some details that should not be visible to users. Even worse, error messages can change with versions of the underlying object storage.
However for multipart upload the basic error messages -- without any parameters from the actual call that we performed -- should be safe: These messages should all come from a similar operation on a block adaptor, rather than something like a KV. (We might get parameters from a non-S3 underlying store, of course -- but error messages will still be better than 500, api error InternalError: We encountered an internal error, please try again.
)
This issue is now marked as stale after 90 days of inactivity, and will be closed soon. To keep it, mark it with the "no stale" label.
What happened?
When I call the
CompleteMultiPartUpload
API incorrectly, it always responds with a vague error message that doesn't help me, such as:But LakeFS actually logs the real error messages:
I think the S3 Gateway API should also respond with a real error messag when encountered an error to help the caller.
lakeFS Version
0.94.1