On 22 Aug 2013, at 01:29, Andrew Phillips <[email protected]> wrote:

> Quick functionality question: what should happen if one of the "delete chunk" 
> calls fails? Should it continue with subsequent chunks?

Good question. I think that it should raise an exception so that the caller 
knows that the delete didn't fully succeed. Whether it should try to delete 
subsequent chunks if a chunk fails depends on the likelihood that deleting 
subsequent chunks will succeed even if a previous chunk failed.

I think that deleting subsequent chunks is only likely to succeed if the 
previous chunk failed due to a transient error (e.g. network glitch, provider 
glitch). In that case retrying the delete of the previous chunk is also likely 
to succeed. Does jclouds as a whole have a stance on retries, i.e. will it 
usually retry an operation if it fails?

Reply via email to