[tahoe-lafs-trac-stream] [Tahoe-LAFS] #3793: allocate_buckets() shouldn't hardcode BucketWriter lifetime as mapping to connections (was: allocate_buckets() should always list all shares in its response)

[Tahoe-LAFS]

#3793: allocate_buckets() shouldn't hardcode BucketWriter lifetime as mapping to
connections
--------------------------+-----------------------------------
     Reporter:  itamarst  |      Owner:  itamarst
         Type:  task      |     Status:  new
     Priority:  normal    |  Milestone:  HTTP Storage Protocol
    Component:  unknown   |    Version:  n/a
   Resolution:            |   Keywords:
Launchpad Bug:            |
--------------------------+-----------------------------------
Description changed by itamarst:

Old description:

> Original issue: Currently, if you allocate buckets twice with the same
> inputs, without writing to any buckets, the result is `(allocated = empty
> set(), writeable buckets = empty dict())`. This is wrong, each share
> number should always be one of the two categories.
>

> Later decided the above is wrong. See below comment 4 for new plan for
> these semantics.

New description:

 Original issue: Currently, if you allocate buckets twice with the same
 inputs, without writing to any buckets, the result is `(allocated = empty
 set(), writeable buckets = empty dict())`. This is wrong, each share
 number should always be one of the two categories.


 Later decided the above is wrong, but the more fundamental point is that
 HTTP has no concept of connections, so we shouldn't tie `BucketWriter`
 lifetime to connections. See below comment 4 for the plan for updating
 these semantics.

--

--
Ticket URL: <https://tahoe-lafs.org/trac/tahoe-lafs/ticket/3793#comment:7>
Tahoe-LAFS <https://Tahoe-LAFS.org>
secure decentralized storage