Şükrü

ZincSearch a lightweight alternative to elasticsearch written in Go.

ZincSearch

zincsearch

if the value is release then gin will run in production mode.

Required for creating volume. Learn more about this configuration: https://docs.railway.com/reference/variables#user-provided-configuration-variables

terms aggregation returns max bucket size

Internal batch size for batching records when bulk indexing is done.

Defaults to "data" folder in current working directory if not provided.

create a keyword field for text field. named field.keyword

Algorithm of compress segment file, default is: zstd, supports: snappy, s2, zstd

Maximum results to be returned from server. Defaults to 100

plugin, GSE which size dict need to load, small or big

plugin, GSE where to load user custom dictionary

plugin, GSE analyzer default enable HMM for search, you can disable it.

plugin, GSE analyzer default load stop dictionary, you can disable it.

Default is false, ZinCsearch uses pyroscope server to profiling

pyroscope identifier id, example: zinc-alex

default pyroscope server is: https://pyroscope.dev.zincsearch.com

Enables prometheus metrics on /metrics endpoint

control parallelsim thread num for shard reads

Send anonymous error reports for debugging

Entry DNS, default is: https://15b6d9b8be824b44896f32b0234c32b7@o1218932.ingest.sentry.io/6360942

We use multiple backend indexes for one big index, limit one shard size default is 1GB

Default is true, Enable swagger api document

Send anonymous telemetry info for improving ZincSearch. enabled or disabled.

ZinCsearch uses REDO log to ensure asynchorous is correct, but redo log do a sync to disk every time, we can disable SYNC to have better performance, but it has a risk duplicated documents with a ZINC_BATCH_SIZE

ZinCsearch uses WAL to ensure no loss data, and asynchorous write to backend index. defaults every 1s sync to storage