Environment Variables
We prefer environment variables for configuration as opposed to configuration files or command line options as its easier to provide environment variables in a containerized environment as well as local setup.
| Environment Variable | Default Value | Mandatory | Description |
|---|---|---|---|
| ZINC_FIRST_ADMIN_USER | - | On first run | First admin user of ZincSearch. Not required after first run of ZincSearch. |
| ZINC_FIRST_ADMIN_PASSWORD | - | On first run | Password for first admin user |
| ZINC_SERVER_ADDRESS | 0.0.0.0 | No | ZincSearch server IP address to bind to |
| ZINC_SERVER_PORT | 4080 | No | ZincSearch server listen http port |
| GIN_MODE | - | No | if the value is release then gin will run in production mode. |
| ZINC_DATA_PATH | ./data | No | Defaults to "data" folder in current working directory if not provided. |
| ZINC_ICE_COMPRESSOR | zstd | No | Algorithm of compress segment file, default is: zstd, supports: snappy, s2, zstd |
| ZINC_MAX_RESULTS | 100 | No | Maximum results to be returned from server. Defaults to 100 |
| ZINC_BATCH_SIZE | 1024 | No | Internal batch size for batching records when bulk indexing is done. |
| ZINC_AGGREGATION_TERMS_SIZE | 1000 | No | terms aggregation returns max bucket size |
| ZINC_SHARD_NUM | 3 | No | We use shards to improve performance |
| ZINC_SHARD_MAX_SIZE | 1073741824 | No | We use multiple backend indexes for one big index, limit one shard size default is 1GB |
| ZINC_READ_GORUTINE_NUM | 10 | No | control parallelsim thread num for shard reads |
| ZINC_TELEMETRY | true | No | Send anonymous telemetry info for improving ZincSearch. enabled or disabled. |
| ZINC_SENTRY | true | No | Send anonymous error reports for debugging |
| ZINC_SENTRY_DSN | - | No | Entry DNS, default is: https://15b6d9b8be824b44896f32b0234c32b7@o1218932.ingest.sentry.io/6360942 |
| ZINC_PROMETHEUS_ENABLE | false | No | Enables prometheus metrics on /metrics endpoint |
| ZINC_PLUGIN_ES_VERSION | - | No | es version, for compatible elasticsearch |
| ZINC_PLUGIN_GSE_ENABLE | false | No | plugin, GSE support Chinese analysis |
| ZINC_PLUGIN_GSE_ENABLE_HMM | true | No | plugin, GSE analyzer default enable HMM for search, you can disable it. |
| ZINC_PLUGIN_GSE_ENABLE_STOP | true | No | plugin, GSE analyzer default load stop dictionary, you can disable it. |
| ZINC_PLUGIN_GSE_DICT_EMBED | small | No | plugin, GSE which size dict need to load, small or big |
| ZINC_PLUGIN_GSE_DICT_PATH | ./plugins/gse/dict | No | plugin, GSE where to load user custom dictionary |
| ZINC_PROFILER | false | No | Default is false, ZinCsearch uses pyroscope server to profiling |
| ZINC_PROFILER_SERVER | - | No | default pyroscope server is: https://pyroscope.dev.zincsearch.com |
| ZINC_PROFILER_API_KEY | - | No | pyroscope server api key |
| ZINC_PROFILER_FRIENDLY_PROFILE_ID | - | No | pyroscope identifier id, example: zinc-alex |
| ZINC_ENABLE_TEXT_KEYWORD_MAPPING | false | No | create a keyword field for text field. named field.keyword |
| ZINC_SWAGGER_ENABLE | true | No | Default is true, Enable swagger api document |
| ZINC_WAL_SYNC_INTERVAL | 1s | No | ZinCsearch uses WAL to ensure no loss data, and asynchorous write to backend index. defaults every 1s sync to storage |
| ZINC_WAL_REDOLOG_NO_SYNC | false | No | 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 |