--- metadata: - name: generator content: Diplodoc Platform v5.50.6 alternate: - https://ytsaurus.tech/docs/en/user-guide/data-processing/chyt/try-chyt.md - https://ytsaurus.tech/docs/ru/user-guide/data-processing/chyt/try-chyt.md --- > **Documentation Index:** Fetch the complete configuration index at https://ytsaurus.tech/docs/en/llms.txt # How to try There are several ways to try CHYT. The simplest way is to run a query via the Query Tracker web interface. You can use a public clique with the `ch_public` alias. This is the main public clique found on every cluster of CHYT's presence. {% note info "Note" %} In YTsaurus 24.1 and lower, a clique alias is specified with an asterisk at the beginning. For example, `*ch_public`. In the current version, `*ch_public` and `ch_public` are the same alias. {% endnote %} {% note warning "Attention" %} All YTsaurus users have access to the public clique, so there is a risk that it may not be available due to queries from other users or be occupied by other people's computations. In this regard, important processes and dashboards must not be based on a public clique. A public clique is essentially a demo version of CHYT. We highly recommend creating a separate clique for your team. {% endnote %} ## HTTP interface { #http } A clique is accessed via heavy YTsaurus HTTP proxies. The most convenient way to access a clique is to send a query to the name of the corresponding cluster and process the redirect. When using the cURL utility, this means that you must specify the `--location-trusted` flag. Below is an example of how to query a clique using the cURL command line utility, knowing the cluster name (`YT_PROXY`), the YTsaurus access token (`YT_TOKEN`), and the clique alias (`CHYT_ALIAS`). ```bash $ curl --location-trusted -H "Authorization: OAuth $YT_TOKEN" "$YT_PROXY/chyt?chyt.clique_alias=$CHYT_ALIAS" -d 'SELECT Avg(a) FROM "//sys/clickhouse/sample_table"' 224.30769230769232 ``` ### How to specify a clique alias { #clique_alias_specification } You can specify a clique alias in the special `chyt.clique_alias` URL parameter (see the example above) or in the `user` field using any of the authorization methods (information about the actual user is contained in the token). For example, `http://cluster.domain.com/chyt?user=my-clique&password=my-token`. {% note warning "Attention" %} Authentication (clique alias + token) via different combinations of parameters isn't allowed. For example, the following query contains an error: `-H 'X-ClickHouse-User: my-clique' 'http://cluster.domain.com/chyt?password=my-token'` (the alias is specified in the `X-ClickHouse-User/Key` headers, and the token is specified in the `user/password` URL parameters) {% endnote %} {% note info "Note" %} In the old version, queries use the `http://$YT_PROXY/query` path. When CHYT is accessed using the deprecated `/query` path, the alias is passed via the `database` url parameter. When CHYT is accessed using the new `/chyt` path, the alias is taken from `user` or the `chyt.clique_alias` URL parameter. {% endnote %} ### Exclusive connection port { #chyt_port } Not all clients support custom paths. Special ports can be configured on the HTTP Proxy to access CHYT. These ports don't require custom paths to be specified. To open standard ports 8123 and 8443, edit the very top of `cluster_v1_local.yaml`: ```yaml apiVersion: cluster.ytsaurus.tech/v1 kind: Ytsaurus metadata: name: minisaurus spec: clusterFeatures: httpProxyHaveChytAddress: true coreImage: ... ``` ## ODBC interface { #odbc } You can use CHYT from Tableau via the [Clickhouse ODBC Driver](https://github.com/ClickHouse/clickhouse-odbc). To do this, you just need to make the ODBC configuration represented below. The `YT_TOKEN` variable denotes the YTsaurus access token (`AQAD-qJ…`), `CHYT_ALIAS` is the alias of the clique (`ch_public`), and `YT_PROXY` is the cluster name. ```bash [ClickHouse] Driver = # For example /usr/local/opt/clickhouse-odbc/lib/libclickhouseodbc.dylib username = $CHYT_ALIAS password = $YT_TOKEN database = YT url = http://$YT_PROXY/chyt port = 80 # trace = 1 # TraceFile=/tmp/odbc.log ``` Below is an example of how to access CHYT via the ODBC interface from the command line using the [isql](https://en.wikipedia.org/wiki/UnixODBC) utility. To do this, write the `~/.odbc.ini` configuration file, install the Clickhouse ODBC driver from the official repository, and use the isql utility. ```bash isql -v clickhouse +---------------------------------------+ | Connected! | | | | sql-statement | | help [tablename] | | quit | | | +---------------------------------------+ SQL> SELECT * FROM "//tmp/sample_table" +--+ | a| +--+ | 2| | 9| | 6| | 4| | 8| | 3| | 1| | 5| | 7| | 0| +--+ SQLRowCount returns 10 10 rows fetched ``` ## JDBC interface { #jdbc } The JDBC driver is configured in a similar way to the ODBC driver, see the example below. The `YT_TOKEN` variable denotes the YTsaurus access token (`AQAD-qJ…`), `CHYT_ALIAS` is the ID/alias of the clique (`ch_public`), and `YT_PROXY` is the cluster name. ```bash JDBC URL = jdbc:clickhouse://{host}:{port}/chyt?chyt.clique_alias={clique_alias} host = $YT_PROXY user = $CHYT_ALIAS password = $YT_TOKEN port = 80 ``` Besides that, you need to change the driver properties: - `check_for_redirects=true`; - `use_path_as_db=false`. The JDBC driver can be used as a basis for using [DBeaver](https://dbeaver.io), an open-source database management program. DBeaver 6.1.5 was used for testing. If the ClickHouse-JDBC driver is not installed automatically, [download](https://github.com/yandex/clickhouse-jdbc) it manually. You must use a driver version higher than 0.1.55. ## CLI and Python API { #cli-and-api } CHYT has the official Python API and the official command line utility. For more information, see [CLI and Python API](https://ytsaurus.tech/docs/en/user-guide/data-processing/chyt/cli-and-api.md).