Not entirely. A hypertable is a postgres table chunked over time. There is the assumption that most data and queries are time-relevant, but also that older data

Not entirely. A hypertable is a postgres table chunked over time. There is the assumption that most data and queries are time-relevant, but also that older data is less relevant than new data. Indexes are per chunk. So if the query analyzer understands you only touch 2023 it can omit looking at any chunk that is from other years and keep those out of memory. Same with the indexes.