This week, the SolrCluster team caught up with Solr committer, Shalin Mangar, for a follow-up discussion to his recent blog post, Call Me Maybe: SolrCloud, Jepsen, and Flaky Networks.
Shalin discusses the CAP theorem (Consistency, Availability, Partition tolerance) and the durability of SolrCloud in the face of network partitions and other service interruptions typical of distributed systems. In particular, Shalin describes his work in testing Solr and ZooKeeper with Jepsen, a framework built in Clojure by Kyle Kingsbury to “[break] distributed systems so you don’t have to”. Tune in to learn about how SolrCloud handles the hazards of distributed computing as a highly scalable, fault-tolerant search and NoSQL engine.
Catch the podcast here.