Experience in creating a Ceph storage with a throughput of tebibytes per second
An engineer from Clyso summarized the experience of creating a storage cluster based on a fault-tolerant distributed system Ceph, achieving bandwidth exceeding tebibyte per second. It is noted that this is the first Ceph-based cluster to reach such a benchmark, but before obtaining the presented result, engineers had to overcome a series of not-so-obvious challenges. For instance, to enhance performance by 10-20%, they had to […]
