Changes between Initial Version and Version 1 of Ticket #1571
- Timestamp:
- 2011-10-25T13:00:47Z (13 years ago)
Legend:
- Unmodified
- Added
- Removed
- Modified
-
Ticket #1571 – Description
initial v1 1 1 The current S3 backend implementation waits until a GET request to S3 for the share has completed before creating a share object that can be read. This ticket is to make it support reading as much data as has been received before the GET is complete. Note that the SFTP frontend does a similar thing in {{{OverwriteableFileConsumer}}}; some of that code or the approach it uses may be reusable. 2 2 3 In the case of mutable shares, the client would need to use multiple {{ remote_readv}}} calls with relatively short lengths. This ticket is not about that, only about whether the S3 backend would support such usage efficiently.3 In the case of mutable shares, the client would need to use multiple {{{remote_readv}}} calls with relatively short lengths. This ticket is not about that, only about whether the S3 backend would support such usage efficiently. 4 4 5 5 This ticket might need to be split for mutable and immutable shares, but I'll leave it as one for the time being.
