Open Stack

Today I'm happy to announce the OpenStack Swift 2.7.0 release.

Barring any major last-minute issues being discovered, our
contribution to the OpenStack Mitaka release includes the Swift 2.6.0
release and this 2.7.0 release.

This release includes some significant improvements that improve the
quality of life for both cluster operators and end users.

As always, you can upgrade to this version of Swift with zero end-user
downtime. I recommend that everyone upgrade.

- Tarball: https://tarballs.openstack.org/swift/swift-2.7.0.tar.gz
- Full Release Notes: https://github.com/openstack/swift/blob/master/CHANGELOG

----------------------------
Highlights from this release
----------------------------

- Fast-POST

"Fast-POST" is the mode where `object_post_as_copy` is set to
`False` in the proxy server config. This mode now allows for
fast, efficient updates of metadata without needing to fully
recopy the contents of the object. While the default still is
`object_post_as_copy` as True, the plan is to change the default
to False and then deprecate post-as-copy functionality in later
releases. Fast-POST now supports container-sync functionality.

The end result with this fast-POST update is that end-user POST
operations are much faster, and you can now use fast-POST with
container sync.

- Concurrent reads

Concurrent reads can significantly lower latency on reads as seen by
the end-user. If Swift is trying to read data off of a slow disk, it
can now opportunistically try another location in the cluster (before
the first operation times out). Swift will serve the request with the
first drive that responds, and the user sees lower time-to-first-byte
times when requesting data.

- An ops runbook

Swift's in-tree docs now include an operations runbook to help you do
the day-to-day of running your cluster. You can find the latest
version at http://swift.openstack.org/ops_runbook/ and a SuperUser
blog post introducing it at http://superuser.openstack.org/articles/help-improve-the-openstack-swift-operations-runbook.

- `handoffs_first` is not a more useful object replication mode

When a drive in the cluster fills up, the top priority is to add
capacity to the cluster and let the replication processes move data to
the newly available storage media. The `handoffs_first` replication
mode exists to prioritize this data movement. This feature has been
improved in this release to make sure that **all** data that needs to
leave the drive is prioritized before normal replication checks.

The end result of this feature update is that operators have a safer
and faster way to add new capacity to a full cluster and get the
cluster healthy again.

- Container sync improved

Container sync has been improved to perform a HEAD on the remote side
of the sync for each object being synced. If the object exists on the
remote side, container-sync will no longer transfer the object, thus
significantly lowering the network requirements to use the feature.

And these are just the highlights! I encourage you to read the full
release notes (linked at the top of this message) and upgrade today.

This release is the work of 52 code contributors, including 16 who
contributed for the first time.

Thank you to all who have contributed to Swift, via code or otherwise.
If you'd like to get more involved, join us in #openstack-swift on
freenode IRC.

--John




-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 801 bytes
Desc: OpenPGP digital signature
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20160325/0cda052e/attachment.pgp>