The big feature that is available for the first time with this beta is fast synchronous replication which can improve synchronous replication speed on the order of 100x.
The MemSQL team would appreciate it if you would try the new synchronous replication and synchronous durability features in a test environment, and give us your feedback.
We refreshed 7.0 beta 1 this afternoon. The “thank you” page after you download has a new build ID. The refresh has many bug fixes and new features. We will be updating the documentation over the next few weeks to cover these features.
We have released another refresh of MemSQL 7.0 beta 1. That means
you’ll get a new build if you download it again. This includes
almost all the features that will be in the 7.0 general availability
release; including fast synchronous replication and durability,
seekable columnstores with HASH indexes, row-level columnstore
locking, SPARSE compression for in-memory rowstore tables, full
automatic statistics for all table types, cross-database views,
incremental backup for columnstore tables, and more.
On Nov. 22nd we released the 7.0 Release Candidate (RC). This supports offline upgrade as well as restoring backups from prior supported releases. For upgrades with Ops from prior releases, a new Ops build is needed; it’s coming soon.
Pull down the version menu near the top in the navbar on the left in the MemSQL documentation and choose version v7.0-Beta to see the latest documentation.
Prior details about 7.0 beta 2 are below.
On September 24th we released 7.0 beta 2. This is like 7.0 beta 1 but with hundreds of bug fixes, and a handful of additional features, including built-in time series functions FIRST(), LAST(), and TIME_BUCKET().