Jewels of Distributed Storage
OpenStack Days NYC, Operators Midcycle and Red Hat Ceph Storage 2.0
Today, while I was enjoying the keynotes of old friends at OpenStack Days New York City, the Ceph team at Red Hat was hard at work releasing RHCS 2.0 — the most significant update to Red Hat Ceph Storage since we acquired Inktank in 2014.

Geographic distribution
Ceph is inherently highly available — you could lose a disk, a host, even a rack of storage with the certainty that your data is safely replicated in at least two other places. Ceph’s synchronous nature is a further guarantee: write operations do not return until all three copies have been committed to disk. Like all engineering decisions, this comes with a tradeoff: Ceph is sensitive to network latency as a result of its conservative design. This makes a stretch cluster configuration, where multiple racks of hardware belonging to the same logical cluster...












