Compare Versions - @clickhouse/client

npm / @clickhouse/client / Compare Versions

View on npm

→ Compare

1.18.2

Improvements

• Added a helping WARN level log message with a suggestion to check the keep_alive configuration if the client receives an ECONNRESET error from the server, which can happen when the server closes idle connections after a certain timeout, and the client tries to reuse such a connection from the pool. This can be especially helpful for new users who might not be aware of this aspect of HTTP connection management. The log message is only emitted if the keep_alive option is enabled in the client configuration, and it includes the server's keep-alive timeout value (if available) to assist with troubleshooting. (#597)

How to reproduce the issue that triggers the log message:

const client = createClient({
  // ...
  keep_alive: {
    enabled: true,
    // ❌ DON'T SET THIS VALUE SO HIGH IN PRODUCTION
    idle_socket_ttl: 1_000_000,
  },
  log: {
    level: ClickHouseLogLevel.WARN, // to see the warning logs
  },
})

for (let i = 0; i < 1000; i++) {
  await client.ping({
    // To use a regular query instead of the /ping endpoint
    // which might be configured differently on the server side
    // and have different timeout settings.
    select: true,
  })

  // Wait long enough to let the server close the idle connection,
  // but not too long to let the client remove it from the pool,
  // in other words try to hit the scenario when the race condition
  // happens between the server closing the connection and the client
  // trying to reuse it.
  await sleep(SERVER_KEEP_ALIVE_TIMEOUT_MS - 100)
}

Example log message:

{
  "message": "Ping: idle socket TTL is greater than server keep-alive timeout, try setting idle socket TTL to a value lower than the server keep-alive timeout to prevent unexpected connection resets, see https://c.house/js_keep_alive_econnreset for more details.",
  "args": {
    "operation": "Ping",
    "connection_id": "8dc1c9bd-7895-49b1-8a95-276470151c65",
    "query_id": "beee95af-2e83-4dcb-8e1e-045bd61f4985",
    "request_id": "8dc1c9bd-7895-49b1-8a95-276470151c65:2",
    "socket_id": "8dc1c9bd-7895-49b1-8a95-276470151c65:1",
    "server_keep_alive_timeout_ms": 10000,
    "idle_socket_ttl": 15000
  },
  "module": "HTTP Adapter"
}

1.18.1

Improvements

• Setting log.level default value to ClickHouseLogLevel.WARN instead of ClickHouseLogLevel.OFF to provide better visibility into potential issues without overwhelming users with too much information by default.

const client = createClient({
  // ...
  log: {
    level: ClickHouseLogLevel.WARN, // default is now ClickHouseLogLevel.WARN instead of ClickHouseLogLevel.OFF
  },
})

• Logging is now lazy, which means that the log messages will only be constructed if the log level is appropriate for the message. This can improve performance in cases where constructing the log message is expensive, and the log level is set to ignore such messages. See ClickHouseLogLevel enum for the complete list of log levels. (#520)

const client = createClient({
  // ...
  log: {
    level: ClickHouseLogLevel.TRACE, // to log everything available down to the network level events
  },
})

• Enhanced the logging of the HTTP request / socket lifecycle with additional trace messages and context such as Connection ID (UUID) and Request ID and Socket ID that embed the connection ID for ease of tracing the logs of a particular request across the connection lifecycle. To enable such logs, set the log.level config option to ClickHouseLogLevel.TRACE. (#567)

[2026-02-25T09:19:13.511Z][TRACE][@clickhouse/client][Connection] Insert: received 'close' event, 'free' listener removed
Arguments: {
  operation: 'Insert',
  connection_id: 'da3c9796-5dc5-46ef-83b0-ed1f4422094c',
  query_id: '9dfda627-39a2-41a6-9fc9-8f8716574826',
  request_id: 'da3c9796-5dc5-46ef-83b0-ed1f4422094c:3',
  socket_id: 'da3c9796-5dc5-46ef-83b0-ed1f4422094c:2',
  event: 'close'
}
[2026-02-25T09:19:13.502Z][TRACE][@clickhouse/client][Connection] Query: reusing socket
Arguments: {
  operation: 'Query',
  connection_id: 'da3c9796-5dc5-46ef-83b0-ed1f4422094c',
  query_id: 'ad0127e8-b1c7-4ed6-9681-c0162f7a0ea9',
  request_id: 'da3c9796-5dc5-46ef-83b0-ed1f4422094c:4',
  socket_id: 'da3c9796-5dc5-46ef-83b0-ed1f4422094c:2',
  usage_count: 1
}

• A step towards structured logging: the client now passes rich context to the logger args parameter (e.g. connection_id, query_id, request_id, socket_id). (#576)

Deprecated API

• The drainStream utility function is now deprecated, as the client will handle draining the stream internally when needed. Use client.command() instead, which will handle draining the stream internally when needed. (#578)

• The sleep utility function is now deprecated, as it is not intended to be used outside of the client implementation. Use setTimeout directly or a more full-featured utility library if you need additional features like cancellation or timers management. (#578)

1.17.0

New features

• Added http_status_code to query, insert, and exec commands (#525, [Kinzeng])
• Fixed ignore_error_response not getting passed when using command (#536, [Kinzeng])

1.16.0

New features

• Added support for the new Disposable API (a.k.a the using keyword) (#500)

async function main() {
  await using client = await client.query(…);

  // some code that can throw
  // but thanks to `using` the client will still get closed

  // client is also automatically closed here by calling [Symbol.disaposeAsync]
}

Without the new using keyword it is required to wrap the code that might leak expensive resources like sockets and big buffers in try / finally

async function main() {
  let client
  try {
    client = await createClient(…);
    // some code that can throw
  } finally {
    if (client) {
      await client.close()
    }
  }
}

1.15.0

New features

• Added support for BigInt values in query parameters. (#487, @dalechyn)

1.14.0

New features

• It is now possible to specify custom parse and stringify functions that will be used instead of the standard JSON.parse and JSON.stringify methods for JSON serialization/deserialization when working with JSON* family formats. See ClickHouseClientConfigOptions.json, and a new custom_json_handling example for more details. (#481, looskie)
• (Node.js only) Added an ignore_error_response param to ClickHouseClient.exec, which allows callers to manually handle request errors on the application side. (#483, Kinzeng)

1.13.0

New features

• Server-side exceptions that occur in the middle of the HTTP stream are now handled correctly. This requires ClickHouse 25.11+. Previous ClickHouse versions are unaffected by this change. (#478)

Improvements

• TupleParam constructor now accepts a readonly array to permit more usages. (#465, Malien)

Bug fixes

• Fixed boolean value formatting in query parameters. Boolean values within Array, Tuple, and Map types are now correctly formatted as TRUE/FALSE instead of 1/0 to ensure proper type compatibility with ClickHouse. (#475, baseballyama)

1.12.1

Improvements

• Improved performance of toSearchParams. (#449, twk)

Other

• Added Node.js 24.x to the CI matrix. Node.js 18.x was removed from the CI due to EOL.

1.12.0

Types

• Add missing allow_experimental_join_condition to ClickHouseSettings typing. (#430, looskie)
• Fixed JSONEachRowWithProgress TypeScript flow after the breaking changes in ClickHouse 25.1. RowOrProgress<T> now has an additional variant: SpecialEventRow<T>. The library now additionally exports the parseError method, and newly added isRow / isException type guards. See the updated JSONEachRowWithProgress example (#443)
• Added missing allow_experimental_variant_type (24.1+), allow_experimental_dynamic_type (24.5+), allow_experimental_json_type (24.8+), enable_json_type (25.3+), enable_time_time64_type (25.6+) to ClickHouseSettings typing. (#445)

Improvements

• Add a warning on a socket closed without fully consuming the stream (e.g., when using query or exec method). (#441)
• (Node.js only) An option to use a simple SELECT query for ping checks instead of /ping endpoint. See the new optional argument to the ClickHouseClient.ping method and PingParams typings. Note that the Web version always used a SELECT query by default, as the /ping endpoint does not support CORS, and that cannot be changed. (#442)

Other

• The project now uses Codecov instead of SonarCloud for code coverage reports. (#444)

1.11.2

A minor release to allow further investigation regarding uncaught error issues with #410.

Types

• Added missing lightweight_deletes_sync typing to ClickHouseSettings (#422, pratimapatel2008)

Improvements (Node.js)

• Added a new configuration option: capture_enhanced_stack_trace; see the JS doc in the Node.js client package. Note that it is disabled by default due to a possible performance impact. (#427)
• Added more try-catch blocks to the Node.js connection layer. (#427)

1.11.1

Bug fixes

• Fixed an issue with URLEncoded special characters in the URL configuration for username or password. (#407)

Improvements

• (Node.js only) Added support for streaming on 32-bit platforms. (#403, shevchenkonik)

1.11.0

New features

• It is now possible to provide custom HTTP headers when calling the query/insert/command/exec methods using the http_headers option. NB: http_headers specified this way will override http_headers set on the client instance level. (#394, @DylanRJohnston)
• (Web only) It is now possible to provide a custom fetch implementation to the client. (#315, @lucacasonato)

1.10.1

Bug fixes

• Fixed NULL parameter binding with Tuple, Array, and Map types. (#374)

Improvements

• ClickHouseSettings typings now include session_timeout and session_check settings. (#370)

1.10.0

New features

• Added support for JWT authentication (ClickHouse Cloud feature) in both Node.js and Web API packages (#270). JWT token can be set via access_token client configuration option.

  const client = createClient({
    // ...
    access_token: '<JWT access token>',
  })
  

  Access token can also be configured via the URL params, e.g., https://host:port?access_token=.... It is also possible to override the access token for a particular request (see BaseQueryParams.auth for more details).

  NB: do not mix access token and username/password credentials in the configuration; the client will throw an error if both are set.

1.9.1

• Fixed an uncaught exception that could happen in case of malformed ClickHouse response when response compression is enabled (#363)

1.9.0

New features

• Added input_format_json_throw_on_bad_escape_sequence to the ClickhouseSettings type. (#355, @emmanuel-bonin)
• The client now exports TupleParam wrapper class, allowing tuples to be properly used as query parameters. Added support for JS Map as a query parameter. (#359)

Improvements

• The client will throw a more informative error if the buffered response is larger than the max allowed string length in V8, which is 2**29 - 24 bytes. (#357)

1.8.1

Bug fixes

• When a custom HTTP agent is used, the HTTP or HTTPS request implementation is now correctly chosen based on the URL protocol. (#352)

1.8.0

New features

• Added support for specifying roles via request query parameters. See this example for more details. (@pulpdrew, #328)

1.7.0

Bug fixes

• (Web only) Fixed an issue where streaming large datasets could provide corrupted results. See #333 (PR) for more details.

New features

• Added JSONEachRowWithProgress format support, ProgressRow interface, and isProgressRow type guard. See this Node.js example for more details. It should work similarly with the Web version.

• (Experimental) Exposed the parseColumnType function that takes a string representation of a ClickHouse type (e.g., FixedString(16), Nullable(Int32), etc.) and returns an AST-like object that represents the type. For example:

  for (const type of [
    'Int32',
    'Array(Nullable(String))',
    `Map(Int32, DateTime64(9, 'UTC'))`,
  ]) {
    console.log(`##### Source ClickHouse type: ${type}`)
    console.log(parseColumnType(type))
  }
  

  The above code will output:

  ##### Source ClickHouse type: Int32
  { type: 'Simple', columnType: 'Int32', sourceType: 'Int32' }
  ##### Source ClickHouse type: Array(Nullable(String))
  {
    type: 'Array',
    value: {
      type: 'Nullable',
      sourceType: 'Nullable(String)',
      value: { type: 'Simple', columnType: 'String', sourceType: 'String' }
    },
    dimensions: 1,
    sourceType: 'Array(Nullable(String))'
  }
  ##### Source ClickHouse type: Map(Int32, DateTime64(9, 'UTC'))
  {
    type: 'Map',
    key: { type: 'Simple', columnType: 'Int32', sourceType: 'Int32' },
    value: {
      type: 'DateTime64',
      timezone: 'UTC',
      precision: 9,
      sourceType: "DateTime64(9, 'UTC')"
    },
    sourceType: "Map(Int32, DateTime64(9, 'UTC'))"
  }
  

  While the original intention was to use this function internally for Native/RowBinaryWithNamesAndTypes data formats headers parsing, it can be useful for other purposes as well (e.g., interfaces generation, or custom JSON serializers).

  NB: currently unsupported source types to parse:

  • Geo
  • (Simple)AggregateFunction
  • Nested
  • Old/new experimental JSON
  • Dynamic
  • Variant

1.6.0

New features

• Added optional real_time_microseconds field to the ClickHouseSummary interface (see https://github.com/ClickHouse/ClickHouse/pull/69032)

Bug fixes

• (Node.js) Fixed unhandled exceptions produced when calling ResultSet.json if the response data was not, in fact, a valid JSON. (#311)

1.5.0

New features

• (Node.js only) It is now possible to disable the automatic decompression of the response stream with the exec method. See ExecParams.decompress_response_stream for more details. (#298).

1.4.1

Improvements

• ClickHouseClient is now exported as a value from @clickhouse/client and @clickhouse/client-web packages, allowing for better integration in dependency injection frameworks that rely on IoC (e.g., Nest.js, tsyringe) (@mathieu-bour, #292).

Bug fixes

• Fixed a potential socket hang-up issue that could happen under 100% CPU load (#294).

1.4.0

New features

• (Node.js only) The exec method now accepts an optional values parameter, which allows you to pass the request body as a Stream.Readable. This can be useful in the case of custom insert streaming with arbitrary ClickHouse data formats (which might not be explicitly supported and allowed by the client in the insert method yet). NB: in this case, you are expected to serialize the data in the stream in the required input format yourself. See #290 for more details.

Improvements

• (Node.js only) The client package now exports a utility method drainStream

1.3.0

New features

• It is now possible to get the entire response headers object from the query/insert/command/exec methods. With query, you can access the ResultSet.response_headers property; other methods (insert/command/exec) return it as parts of their response objects as well. For example:

  const rs = await client.query({
    query: 'SELECT * FROM system.numbers LIMIT 1',
    format: 'JSONEachRow',
  })
  console.log(rs.response_headers['content-type'])
  

  This will print: application/x-ndjson; charset=UTF-8. It can be used in a similar way with the other methods.

Improvements

• Re-exported several constants from the @clickhouse/client-common package for convenience:

  • SupportedJSONFormats
  • SupportedRawFormats
  • StreamableFormats
  • StreamableJSONFormats
  • SingleDocumentJSONFormats
  • RecordsJSONFormats

1.2.0

New features

• (Experimental) Added an option to provide a custom HTTP Agent in the client configuration via the http_agent option (#283, related: #278). The following conditions apply if a custom HTTP Agent is provided:

  • The max_open_connections and tls options will have no effect and will be ignored by the client, as those are part of the underlying HTTP Agent configuration.
  • keep_alive.enabled will only regulate the default value of the Connection header (true -> Connection: keep-alive, false -> Connection: close).
  • While the idle socket management will still work, it is now possible to disable it completely by setting the keep_alive.idle_socket_ttl value to 0.

• (Experimental) Added a new client configuration option: set_basic_auth_header, which controls whether the Authorization header should be set for every outgoing HTTP request (enabled by default). One of the possible scenarios when it is necessary to disable this header is when a custom HTTPS agent is used, and the server requires TLS with certificates. For example:

  const agent = new https.Agent({
    keepAlive: true,
    keepAliveMsecs: 2500,
    maxSockets: 10,
    maxFreeSockets: 10,
    ca: fs.readFileSync('./ca.crt'),
    cert: fs.readFileSync('./client.crt'),
    key: fs.readFileSync('./client.key'),
  })
  const client = createClient({
    url: 'https://myserver:8443',
    http_agent: agent,
    // With a custom HTTPS agent, the client won't use the default HTTPS connection implementation; the headers should be provided manually
    http_headers: {
      'X-ClickHouse-User': 'username',
      'X-ClickHouse-Key': 'password',
      'X-ClickHouse-SSL-Certificate-Auth': 'on',
    },
    // Important: authorization header conflicts with the TLS headers; disable it.
    set_basic_auth_header: false,
  })
  

NB: It is currently not possible to set the set_basic_auth_header option via the URL params.

See the doc entry regarding custom HTTP(s) Agent usage with code samples.

If you have feedback on these experimental features, please let us know by creating an issue in the repository or send a message in the Community Slack (#clickhouse-js channel).

1.1.0

New features

• Added an option to override the credentials for a particular query/command/exec/insert request via the BaseQueryParams.auth setting; when set, the credentials will be taken from there instead of the username/password provided during the client instantiation (#278).
• Added an option to override the session_id for a particular query/command/exec/insert request via the BaseQueryParams.session_id setting; when set, it will be used instead of the session id provided during the client instantiation (@holi0317, #271).

Bug fixes

• Fixed the incorrect ResponseJSON<T>.totals TypeScript type. Now it correctly matches the shape of the data (T, default = unknown) instead of the former Record<string, number> definition (#274).

1.0.2

Bug fixes

• The command method now drains the response stream properly, as the previous implementation could cause the Keep-Alive socket to close after each request.
• (Node.js) Removed an unnecessary error log in the ResultSet.stream method if the request was aborted or the result set was closed (#263).

Improvements

• (Node.js) ResultSet.stream logs an error via the Logger instance if the stream emits an error event instead of a simple console.error call.
• Minor adjustments to the formatting of the DefaultLogger log messages.
• Added missing rows_before_limit_at_least to the ResponseJSON type (@0237h, #267).

1.0.1

A hotfix for both Node.js and Web packages. See the 1.0.0 entry that contains the entire 1.0.x changelog.

Bug fixes

• Fixed the regression where the default HTTP/HTTPS port numbers (80/443) could not be used with the URL configuration (#258).

1.0.0

Formal stable release milestone with many improvements and some breaking changes.

Major new features overview:

• Advanced TypeScript support for query + ResultSet
• URL configuration

From now on, the client will follow the official semantic versioning guidelines.

Deprecated API

The following configuration parameters are marked as deprecated:

• The host configuration parameter is deprecated; use url instead.
• additional_headers configuration parameter is deprecated; use http_headers instead.

The client will log a warning if any of these parameters are used. However, it is still allowed to use host instead of url and additional_headers instead of http_headers for now; this deprecation is not supposed to break the existing code.

These parameters will be removed in the next major release (2.0.0).

See the "New features" section for more details.

Breaking changes in 1.0.0

• compression.response is now disabled by default in the client configuration options, as it cannot be used with readonly=1 users, and it was not clear from the ClickHouse error message what exact client option was causing the failing query in this case. If you'd like to continue using response compression, you should explicitly enable it in the client configuration.
• As the client now supports parsing URL configuration, you should specify pathname as a separate configuration option (as it would be considered as the database otherwise).
• (TypeScript only) ResultSet and Row are now more strictly typed, according to the format used during the query call.
• (TypeScript only) Both Node.js and Web versions now uniformly export correct ClickHouseClient and ClickHouseClientConfigOptions types specific to each implementation. Exported ClickHouseClient now does not have a Stream type parameter, as it was unintended to expose it there. NB: you should still use the createClient factory function provided in the package.

New features in 1.0.0

Advanced TypeScript support for query + ResultSet

The client will now try its best to figure out the shape of the data based on the DataFormat literal specified to the query call, as well as which methods are allowed to be called on the ResultSet.

Live demo (see the full description below):

Screencast

Complete reference:

Format	ResultSet.json<T>()	ResultSet.stream<T>()	Stream data	Row.json<T>()
JSON	ResponseJSON<T>	never	never	never
JSONObjectEachRow	Record<string, T>	never	never	never
All other JSON*EachRow	Array<T>	Stream<Array<Row<T>>>	Array<Row<T>>	T
CSV/TSV/CustomSeparated/Parquet	never	Stream<Array<Row<T>>>	Array<Row<T>>	never

By default, T (which represents JSONType) is still unknown. However, considering the JSONObjectsEachRow example, prior to 1.0.0, you had to specify the entire type hint, including the shape of the data, manually:

type Data = { foo: string }

const resultSet = await client.query({
  query: 'SELECT * FROM my_table',
  format: 'JSONObjectsEachRow',
})

// pre-1.0.0, `resultOld` has type Record<string, Data>
const resultOld = resultSet.json<Record<string, Data>>()
// const resultOld = resultSet.json<Data>() // incorrect! The type hint should've been `Record<string, Data>` here.

// 1.0.0, `resultNew` also has type Record<string, Data>; the client inferred that it has to be a Record from the format literal.
const resultNew = resultSet.json<Data>()

This is even more handy in the case of streaming on the Node.js platform:

const resultSet = await client.query({
  query: 'SELECT * FROM my_table',
  format: 'JSONEachRow',
})

// pre-1.0.0
// `streamOld` was just a regular Node.js Stream.Readable
const streamOld = resultSet.stream()
// `rows` were `any`, needed an explicit type hint
streamNew.on('data', (rows: Row[]) => {
  rows.forEach((row) => {
    // without an explicit type hint to `rows`, calling `forEach` and other array methods resulted in TS compiler errors
    const t = row.text
    const j = row.json<Data>() // `j` needed a type hint here, otherwise, it's `unknown`
  })
})

// 1.0.0
// `streamNew` is now StreamReadable<T> (Node.js Stream.Readable with a bit more type hints);
// type hint for the further `json` calls can be added here (and removed from the `json` calls)
const streamNew = resultSet.stream<Data>()
// `rows` are inferred as an Array<Row<Data, "JSONEachRow">> instead of `any`
streamNew.on('data', (rows) => {
  // `row` is inferred as Row<Data, "JSONEachRow">
  rows.forEach((row) => {
    // no explicit type hints required, you can use `forEach` straight away and TS compiler will be happy
    const t = row.text
    const j = row.json() // `j` will be of type Data
  })
})

// async iterator now also has type hints
//Similarly to the `on(data)` example above, `rows` are inferred as Array<Row<Data, "JSONEachRow">>
for await (const rows of streamNew) {
  // `row` is inferred as Row<Data, "JSONEachRow">
  rows.forEach((row) => {
    const t = row.text
    const j = row.json() // `j` will be of type Data
  })
}

Calling ResultSet.stream is not allowed for specific data formats, such as JSON and JSONObjectsEachRow (unlike JSONEachRow and the rest of JSON*EachRow, these formats return a single object). In these cases, the client throws an error. However, it was previously not reflected on the type level; now, calling stream on these formats will result in a TS compiler error. For example:

const resultSet = await client.query('SELECT * FROM table', {
  format: 'JSON',
})
const stream = resultSet.stream() // `stream` is `never`

Calling ResultSet.json also does not make sense on CSV and similar "raw" formats, and the client throws. Again, now, it is typed properly:

const resultSet = await client.query('SELECT * FROM table', {
  format: 'CSV',
})
// `json` is `never`; same if you stream CSV, and call `Row.json` - it will be `never`, too.
const json = resultSet.json()

Currently, there is one known limitation: as the general shape of the data and the methods allowed for calling are inferred from the format literal, there might be situations where it will fail to do so, for example:

// assuming that `queryParams` has `JSONObjectsEachRow` format inside
async function runQuery(
  queryParams: QueryParams,
): Promise<Record<string, Data>> {
  const resultSet = await client.query(queryParams)
  // type hint here will provide a union of all known shapes instead of a specific one
  // inferred shapes: Data[] | ResponseJSON<Data> | Record<string, Data>
  return resultSet.json<Data>()
}

In this case, as it is likely that you already know the desired format in advance (otherwise, returning a specific shape like Record<string, Data> would've been incorrect), consider helping the client a bit:

async function runQuery(
  queryParams: QueryParams,
): Promise<Record<string, Data>> {
  const resultSet = await client.query({
    ...queryParams,
    format: 'JSONObjectsEachRow',
  })
  // TS understands that it is a Record<string, Data> now
  return resultSet.json<Data>()
}

If you are interested in more details, see the related test (featuring a great ESLint plugin expect-types) in the client package.

URL configuration

• Added url configuration parameter. It is intended to replace the deprecated host, which was already supposed to be passed as a valid URL.
• It is now possible to configure most of the client instance parameters with a URL. The URL format is http[s]://[username:password@]hostname:port[/database][?param1=value1&param2=value2]. In almost every case, the name of a particular parameter reflects its path in the config options interface, with a few exceptions. The following parameters are supported:

Parameter	Type
pathname	an arbitrary string.
application_id	an arbitrary string.
session_id	an arbitrary string.
request_timeout	non-negative number.
max_open_connections	non-negative number, greater than zero.
compression_request	boolean. See below [1].
compression_response	boolean.
log_level	allowed values: OFF, TRACE, DEBUG, INFO, WARN, ERROR.
keep_alive_enabled	boolean.
clickhouse_setting_* or ch_*	see below [2].
http_header_*	see below [3].
(Node.js only) keep_alive_idle_socket_ttl	non-negative number.

[1] For booleans, valid values will be true/1 and false/0.

[2] Any parameter prefixed with clickhouse_setting_ or ch_ will have this prefix removed and the rest added to client's clickhouse_settings. For example, ?ch_async_insert=1&ch_wait_for_async_insert=1 will be the same as:

createClient({
  clickhouse_settings: {
    async_insert: 1,
    wait_for_async_insert: 1,
  },
})

Note: boolean values for clickhouse_settings should be passed as 1/0 in the URL.

[3] Similar to [2], but for the http_header configuration. For example, ?http_header_x-clickhouse-auth=foobar will be an equivalent of:

createClient({
  http_headers: {
    'x-clickhouse-auth': 'foobar',
  },
})

Important: URL will always overwrite the hardcoded values and a warning will be logged in this case.

Currently not supported via URL:

• log.LoggerClass
• (Node.js only) tls_ca_cert, tls_cert, tls_key.

See also: URL configuration example.

Performance

• (Node.js only) Improved performance when decoding the entire set of rows with streamable JSON formats (such as JSONEachRow or JSONCompactEachRow) by calling the ResultSet.json() method. NB: The actual streaming performance when consuming the ResultSet.stream() hasn't changed. Only the ResultSet.json() method used a suboptimal stream processing in some instances, and now ResultSet.json() just consumes the same stream transformer provided by the ResultSet.stream() method (see #253 for more details).

Miscellaneous

• Added http_headers configuration parameter as a direct replacement for additional_headers. Functionally, it is the same, and the change is purely cosmetic, as we'd like to leave an option to implement a TCP connection open in the future.

0.3.1

Bug fixes

• Fixed an issue where query parameters containing tabs or newline characters were not encoded properly (#249).

0.3.0

This release primarily focuses on improving the Keep-Alive mechanism's reliability on the client side.

New features

• Idle sockets timeout rework; now, the client attaches internal timers to idling sockets and forcefully removes them from the pool if it is considered that a particular socket has been idling for too long. This additional socket housekeeping intends to eliminate "Socket hang-up" errors that could previously still occur on certain configurations. Now, the client does not rely on the KeepAlive agent when it comes to removing the idling sockets; in most cases, the server will not close the socket before the client does.

• There is a new keep_alive.idle_socket_ttl configuration parameter. The default value is 2500 (milliseconds), which is considered to be safe, as ClickHouse versions prior to 23.11 had keep_alive_timeout set to 3 seconds by default, and keep_alive.idle_socket_ttl is supposed to be slightly less than that to allow the client to remove the sockets that are about to expire before the server does so.

• Logging improvements: more internal logs on failing requests; all client methods except ping will log an error on failure now. A failed ping will log a warning since the underlying error is returned as a part of its result. Client logging still needs to be enabled explicitly by specifying the desired log.level config option, as the log level is OFF by default. Currently, the client logs the following events, depending on the selected log.level value:

  • TRACE - low-level information about the Keep-Alive sockets lifecycle.
  • DEBUG - response information (without authorization headers and host info).
  • INFO - still mostly unused, will print the current log level when the client is initialized.
  • WARN - non-fatal errors; failed ping request is logged as a warning, as the underlying error is included in the returned result.
  • ERROR - fatal errors from query/insert/exec/command methods, such as a failed request.

Breaking changes

• keep_alive.retry_on_expired_socket and keep_alive.socket_ttl configuration parameters are removed.
• The max_open_connections configuration parameter is now 10 by default, as we should not rely on the KeepAlive agent's defaults.
• Fixed the default request_timeout configuration value (now it is correctly set to 30_000, previously 300_000 (milliseconds)).

Bug fixes

• Fixed a bug with Ping that could lead to an unhandled "Socket hang-up" propagation.
• The client ensures proper Connection header value considering Keep-Alive settings. If Keep-Alive is disabled, its value is now forced to "close".

0.2.10

New features

• If InsertParams.values is an empty array, no request is sent to the server and ClickHouseClient.insert short-circuits itself. In this scenario, the newly added InsertResult.executed flag will be false, and InsertResult.query_id will be an empty string.

Bug fixes

• Client no longer produces Code: 354. inflate failed: buffer error exception if request compression is enabled and InsertParams.values is an empty array (see above).

0.2.9

New features

• It is now possible to set additional HTTP headers for outgoing ClickHouse requests. This might be useful if, for example, you have a reverse proxy with authorization. (@teawithfruit)

const client = createClient({
  additional_headers: {
    'X-ClickHouse-User': 'clickhouse_user',
    'X-ClickHouse-Key': 'clickhouse_password',
  },
})

0.2.8

New features

• (Web only) Allow to modify Keep-Alive setting (previously always disabled). Keep-Alive setting is now enabled by default for the Web version.

import { createClient } from '@clickhouse/client-web'
const client = createClient({ keep_alive: { enabled: true } })

• (Node.js & Web) It is now possible to either specify a list of columns to insert the data into or a list of excluded columns:

// Generated query: INSERT INTO mytable (message) FORMAT JSONEachRow
await client.insert({
  table: 'mytable',
  format: 'JSONEachRow',
  values: [{ message: 'foo' }],
  columns: ['message'],
})

// Generated query: INSERT INTO mytable (* EXCEPT (message)) FORMAT JSONEachRow
await client.insert({
  table: 'mytable',
  format: 'JSONEachRow',
  values: [{ id: 42 }],
  columns: { except: ['message'] },
})

See also the new examples:

• Including specific columns or excluding certain ones instead
• Leveraging this feature when working with ephemeral columns (#217)

0.2.7

New features

• (Node.js only) X-ClickHouse-Summary response header is now parsed when working with insert/exec/command methods. See the related test for more details. NB: it is guaranteed to be correct only for non-streaming scenarios. The web version does not currently support this due to CORS limitations. (#210)

Bug fixes

• Drain insert response stream in Web version - required to properly work with async_insert, especially in the Cloudflare Workers context.

0.2.6

New features

• Added Parquet format streaming support to the Node.js client. Examples: insert from a file, select into a file.

0.2.5

Bug fixes

• pathname segment from host client configuration parameter is now handled properly when making requests. See this comment for more details.

0.2.4

No changes in web/common modules.

Bug fixes

• (Node.js only) Fixed an issue where streaming large datasets could provide corrupted results. See #171 (issue) and #204 (PR) for more details.

0.2.3

No changes in web/common modules.

Bug fixes

• (Node.js only) Fixed an issue where the underlying socket was closed every time after using insert with a keep_alive option enabled, which led to performance limitations. See #202 for more details. (@varrocs)

0.2.2

New features

• Added missing default_format setting, which allows to perform exec calls without the FORMAT clause. See the example.

0.2.1

Breaking changes

Date objects in query parameters are now serialized as time-zone-agnostic Unix timestamps (NNNNNNNNNN[.NNN], optionally with millisecond-precision) instead of datetime strings without time zones (YYYY-MM-DD HH:MM:SS[.MMM]). This means the server will receive the same absolute timestamp the client sent even if the client's time zone and the database server's time zone differ. Previously, if the server used one time zone and the client used another, Date objects would be encoded in the client's time zone and decoded in the server's time zone and create a mismatch.

For instance, if the server used UTC (GMT) and the client used PST (GMT-8), a Date object for "2023-01-01 13:00:00 PST" would be encoded as "2023-01-01 13:00:00.000" and decoded as "2023-01-01 13:00:00 UTC" (which is 2023-01-01 05:00:00 PST). Now, "2023-01-01 13:00:00 PST" is encoded as "1672606800000" and decoded as "2023-01-01 21:00:00 UTC", the same time the client sent.

Props to @ide for implementing it.

0.2.0

Introduces web client (using native fetch and WebStream APIs) without Node.js modules in the common interfaces. No polyfills are required.

The web client is confirmed to work with Chrome/Firefox/CloudFlare workers.

It is now possible to implement new custom connections on top of @clickhouse/client-common.

The repository was refactored into three packages:

• @clickhouse/client-common: all possible platform-independent code, types and interfaces
• @clickhouse/client-web: new web (or non-Node.js env) connection, uses native fetch.
• @clickhouse/client: Node.js connection as it was before.

Node.js client breaking changes

• Changed ping method behavior: it will not throw now. Instead, either { success: true } or { success: false, error: Error } is returned.
• Log level configuration parameter is now explicit instead of CLICKHOUSE_LOG_LEVEL environment variable. Default is OFF.
• query return type signature changed to is BaseResultSet<Stream.Readable> (no functional changes)
• exec return type signature changed to ExecResult<Stream.Readable> (no functional changes)
• insert<T> params argument type changed to InsertParams<Stream, T> (no functional changes)
• Experimental schema module is removed

Web client known limitations

• Streaming for select queries works, but it is disabled for inserts (on the type level as well).
• KeepAlive is disabled and not configurable yet.
• Request compression is disabled and configuration is ignored. Response compression works.
• No logging support yet.

0.1.1

New features

• Expired socket detection on the client side when using Keep-Alive. If a potentially expired socket is detected, and retry is enabled in the configuration, both socket and request will be immediately destroyed (before sending the data), and the client will recreate the request. See ClickHouseClientConfigOptions.keep_alive for more details. Disabled by default.
• Allow disabling Keep-Alive feature entirely.
• TRACE log level.

Examples

Disable Keep-Alive feature

const client = createClient({
  keep_alive: {
    enabled: false,
  },
})

Retry on expired socket

const client = createClient({
  keep_alive: {
    enabled: true,
    // should be slightly less than the `keep_alive_timeout` setting in server's `config.xml`
    // default is 3s there, so 2500 milliseconds seems to be a safe client value in this scenario
    // another example: if your configuration has `keep_alive_timeout` set to 60s, you could put 59_000 here
    socket_ttl: 2500,
    retry_on_expired_socket: true,
  },
})

0.1.0

Breaking changes

• connect_timeout client setting is removed, as it was unused in the code.

New features

• command method is introduced as an alternative to exec. command does not expect user to consume the response stream, and it is destroyed immediately. Essentially, this is a shortcut to exec that destroys the stream under the hood. Consider using command instead of exec for DDLs and other custom commands which do not provide any valuable output.

Example:

// incorrect: stream is not consumed and not destroyed, request will be timed out eventually
await client.exec('CREATE TABLE foo (id String) ENGINE Memory')

// correct: stream does not contain any information and just destroyed
const { stream } = await client.exec('CREATE TABLE foo (id String) ENGINE Memory')
stream.destroy()

// correct: same as exec + stream.destroy()
await client.command('CREATE TABLE foo (id String) ENGINE Memory')

Bug fixes

• Fixed delays on subsequent requests after calling insert that happened due to unclosed stream instance when using low number of max_open_connections. See #161 for more details.
• Request timeouts internal logic rework (see #168)

0.0.16

Breaking changes

• Node.js 14 EOL as its maintenance phase has ended in April 2023. Node.js 16+ is now required to use the client.

Bug fixes

• Fix NULL parameter binding. As the HTTP interface expects \N instead of a 'NULL' string, it is now correctly handled for both null and explicitly undefined parameters. See the test scenarios for more details.

0.0.15

Bug fixes

• Fix Node.JS 19.x/20.x timeout error (@olexiyb)

0.0.14

New features

• Added support for JSONStrings, JSONCompact, JSONCompactStrings, JSONColumnsWithMetadata formats (@andrewzolotukhin).

0.0.13

New features

• query_id can now be overridden for all main client's methods: query, exec, insert.

0.0.12

New features

• ResultSet.query_id contains a unique query identifier that might be useful for retrieving query metrics from system.query_log
• User-Agent HTTP header is set according to the language client spec. For example, for client version 0.0.12 and Node.js runtime v19.0.4 on Linux platform, it will be clickhouse-js/0.0.12 (lv:nodejs/19.0.4; os:linux). If ClickHouseClientConfigOptions.application is set, it will be prepended to the generated User-Agent.
• Run tests on nodejs@v19

Breaking changes

• client.insert now returns { query_id: string } instead of void
• client.exec now returns { stream: Stream.Readable, query_id: string } instead of just Stream.Readable

0.0.11

Breaking changes

• log.enabled flag was removed from the client configuration.
• Use the CLICKHOUSE_LOG_LEVEL environment variable instead. Possible values: OFF, TRACE, DEBUG, INFO, WARN, ERROR. Currently, there are only debug messages, but we will log more later.

For more details, see PR #110

0.0.10

Remove request listeners synchronously (#124) - closed issue #123

0.0.9

See #121

import { createClient } from '@clickhouse/client'

const client = createClient({
  session_id: `<session_id>`,
})

Kudos to @KMahoney

0.0.8

Added SSL/TLS support (basic and mutual), resolving #52

0.0.7

Resolved #116

0.0.6

See #113

0.0.5

• Breaking changes: Rows abstraction was renamed to ResultSet.
• Breaking changes: now, every iteration over ResultSet.stream() yields Row[] instead of a single Row. Please check out an example and this PR for more details. These changes allowed us to significantly reduce overhead on select result set streaming.
• Updated README and examples to cover the changes.
• split2 is no longer a package dependency.

0.0.4

Remove package.json dependency