Contents

Idea Repository

Contents

List of ideas to explore.

  • Interaction between a filesystem and a computational SSD that does transparent compression.

    • Problem: How to manage SSD space in this case? What API should be use?
    • Can use SSDSim to explore the design space.
  • Key prefix compression in k/v store

    • Problem: in levelDB, it uses fixed-size key prefix compression. Can we do better than this?
    • In Rockset/Foundationdb, secondary indexes are stored as <key, null> key/value pairs. This creates an opportunity that significant storage saving can be achieved if we can do a better key compression.
  • Zoned namespace SSDs

  • VectorBtree, VectorBloomFilter, VectorKV, VectorDB, VectorFS Investigate how to levarage AVX extension instructions to accelerate a B/B+ tree index, certain operations in a KV store, a DB or a FS.

    Execution plan:

    • Understand how lookup works in an indexing data structure; determine whether we can easily do batch lookups;
    • find an open-source implementation; modify and add support for batch lookup;
    • measure the batch lookup speedup compared with point lookup -> best result.
    • Integrate batch lookup into some systems, such as a KV store or a DB.
  • VectorZip
    Use AVX instruction to accelerate compression (zstd/lz4)