Cassandra Admin case studies
Cassandra Admin · Put your knowledge to work in context.
The Cassandra Admin credential certifies that a candidate can run Apache Cassandra in production rather than merely write CQL against it. It targets database administrators, SREs and platform engineers who already touch a cluster and want their operational judgment validated: reading nodetool output under pressure, sizing a compaction strategy, deciding when a repair window is safe, and knowing which consistency level actually protects a given write. The Cassandra Admin case studies are 5 cases and about 4,706 words.
1Meter readings: an unbounded partition that keeps growing for 24 months
An energy utility stores smart-meter readings. 12,000 meters report every 30 seconds - roughly 34 million writes per day, retained for 24 months.
922 words, about 5 min2Session-state table: compaction backlog, oversized heap, and 34 SSTables per read
A six-node single-datacenter cluster (RF=3, 32 GB RAM, 16 cores, NVMe, 3 TB per node) holds a session-state workload. Writes are 90% overwrites of a key that was written less than five minutes earlier; rows carry a one-hour TTL and are read only by primary key.
954 words, about 5 min3Two-DC cluster: a restore that resurrected 4.1 million purged rows
A twelve-node cluster spans two datacenters: us-east with eight nodes and eu-west with four, both at RF=3, both receiving all writes at LOCAL QUORUM. nodetool repair -pr runs from cron on each node once every 30 days, sequentially, at about 14 hours per node.
945 words, about 5 min4Enabling authentication, TLS and audit on a shared unauthenticated cluster
A nine-node, single-datacenter cluster (RF=3) serves three internal teams with a keyspace-per-team layout. Today it runs AllowAllAuthenticator and AllowAllAuthorizer , has no internode encryption, and clients connect over the public interface because the private network was never provisioned.
943 words, about 5 min54.0 to 4.1 upgrade on a cluster that is already dropping mutations
A fifteen-node single-datacenter cluster (RF=3) on Cassandra 4.0.7, with 6 TB per node and all user tables on SizeTieredCompactionStrategy, is scheduled to move to 4.1.3. It backs a 24/7 API with a 99.95% availability target and a p99 read of 12 ms.
942 words, about 5 min