Choosing a RocksDB Compaction Style for Write-Heavy Embedded Workloads
A decision guide for picking between RocksDB's leveled, universal, and FIFO compaction styles for write-heavy embedded workloads, with constraints and validation.
Loading…
ReadMeFeed / Community knowledge
#compaction
Real questions. Useful conversations. Find the people who know your stack.
RocksDB Column Families: One Database, Independently Tuned Data Sets
RocksDB column families let you give hot sessions and cold audit logs different compaction and compression policies inside one database. A worked C++ example plus verification and pitfalls.
Diagnosing RocksDB Write Stalls: From 'Stalling Writes' to a Targeted Fix
RocksDB write stalls come from exactly three backlogs: unflushed memtables, L0 file accumulation, or pending compaction bytes. Here's how to tell them apart and fix the right one.
Cassandra Compaction for Time-Series: Why TWCS Wins Until Writes Aren't Ordered
Compaction strategy is a data-lifecycle decision, not a tuning knob. How TWCS windows work for time-series, where out-of-order writes break them, and how to verify.
cassandraChoosing a RocksDB Compaction Style for Your Workload
Guide to picking Level, Universal, or FIFO compaction in RocksDB based on read/write ratio, latency, and storage constraints, with sample C++/Java code and verification steps.
rocksdbUsing Cassandra TTL to Automate Stale‑Data Cleanup – What You Need to Know
Learn how Cassandra’s built‑in TTL feature can automatically expire data, why it creates tombstones, and how to monitor and tune it to avoid performance surprises.
Selecting STCS or LCS for Write‑Intensive Cassandra Tables
Learn how to choose between SizeTiered and Leveled compaction in Cassandra for write‑heavy tables, with a concrete example of switching strategies and verification steps.
Isolate Workloads in RocksDB Using Column Families with Custom Memtable and Compaction Settings
Learn how to configure RocksDB column families to give different workloads their own memtable, write buffer, and compaction settings while sharing a single DB instance.
TTL expiration behavior with merge operators in RocksDB compaction
Goal Determine whether RocksDB removes TTL‑expired keys before applying a merge operator during compaction, and what impact this interaction has on write amplification and memtable usage. Constraints / Uncertainty The documentation states that TTL expiration is performed lazily during compaction or reads, but does not specify the order relative to merge‑oper