ReductStore is a time-series database designed specifically for blob (binary) data.
ReductStore packs blobs that are smaller than the minimum size of file system blocks into larger logical blocks that are stored as files to reduce wasted disk space. No compression algorithm is applied at any level.
Concurrency control is not supported. Queries are processed strictly in the order of their arrival time in a FIFO manner.
Each entry is a blob associated with a key at creation time. Entries smaller than the file system's minimum file size are grouped and stored together in a logical block, which is backed by a single file. A logical bucket contains multiple logical blocks. Bucket ensures storage limits (total blob size and the total number of records) and write behavior when the storage limit is reached (reject write or delete earliest).
Deleting a single entry is not supported. The minimum deletion level is a single bucket.