- **[ClickHouse Cloud](https://clickhouse.com/cloud/):** the official ClickHouse as a service, - built by, maintained, and supported by the creators of ClickHouse
- **[Self-managed ClickHouse](https://github.com/ClickHouse/ClickHouse):** ClickHouse can run on any Linux, FreeBSD, or Mac OS X with x86_64, AArch64, or PowerPC64LE CPU architecture
The quickest and easiest way to get up and running with ClickHouse is to create a new service in [ClickHouse Cloud](https://clickhouse.cloud/):
<divclass="eighty-percent">
</div>
Once your Cloud service is provisioned, you will be able to [connect to it](/docs/en/integrations/connect-a-client.md) and start [inserting data](/docs/en/integrations/data-ingestion.md).
:::note
The [Quick Start](/docs/en/quick-start.mdx) walks through the steps to get a ClickHouse Cloud service up and running, connecting to it, and inserting data.
Official pre-built binaries are typically compiled for x86_64 and leverage SSE 4.2 instruction set, so unless otherwise stated usage of CPU that supports it becomes an additional system requirement. Here’s the command to check if current CPU has support for SSE 4.2:
To run ClickHouse on processors that do not support SSE 4.2 or have AArch64 or PowerPC64LE architecture, you should [build ClickHouse from sources](#from-sources) with proper configuration adjustments.
ClickHouse implements parallel data processing and uses all the hardware resources available. When choosing a processor, take into account that ClickHouse works more efficiently at configurations with a large number of cores but a lower clock rate than at configurations with fewer cores and a higher clock rate. For example, 16 cores with 2600 MHz is preferable to 8 cores with 3600 MHz.
It is recommended to use **Turbo Boost** and **hyper-threading** technologies. It significantly improves performance with a typical workload.
### RAM {#ram}
We recommend using a minimum of 4GB of RAM to perform non-trivial queries. The ClickHouse server can run with a much smaller amount of RAM, but it requires memory for processing queries.
The required volume of RAM depends on:
- The complexity of queries.
- The amount of data that is processed in queries.
To calculate the required volume of RAM, you should estimate the size of temporary data for [GROUP BY](/docs/en/sql-reference/statements/select/group-by.md#select-group-by-clause), [DISTINCT](/docs/en/sql-reference/statements/select/distinct.md#select-distinct), [JOIN](/docs/en/sql-reference/statements/select/join.md#select-join) and other operations you use.
ClickHouse can use external memory for temporary data. See [GROUP BY in External Memory](/docs/en/sql-reference/statements/select/group-by.md#select-group-by-in-external-memory) for details.
### Swap File {#swap-file}
Disable the swap file for production environments.
### Storage Subsystem {#storage-subsystem}
You need to have 2GB of free disk space to install ClickHouse.
The volume of storage required for your data should be calculated separately. Assessment should include:
- Estimation of the data volume.
You can take a sample of the data and get the average size of a row from it. Then multiply the value by the number of rows you plan to store.
- The data compression coefficient.
To estimate the data compression coefficient, load a sample of your data into ClickHouse, and compare the actual size of the data with the size of the table stored. For example, clickstream data is usually compressed by 6-10 times.
To calculate the final volume of data to be stored, apply the compression coefficient to the estimated data volume. If you plan to store data in several replicas, then multiply the estimated volume by the number of replicas.
### Network {#network}
If possible, use networks of 10G or higher class.
The network bandwidth is critical for processing distributed queries with a large amount of intermediate data. Besides, network speed affects replication processes.
### Software {#software}
ClickHouse is developed primarily for the Linux family of operating systems. The recommended Linux distribution is Ubuntu. The `tzdata` package should be installed in the system.
It is recommended to use official pre-compiled `tgz` archives for all Linux distributions, where installation of `deb` or `rpm` packages is not possible.
For production environments, it’s recommended to use the latest `stable`-version. You can find its number on GitHub page https://github.com/ClickHouse/ClickHouse/tags with postfix `-stable`.
To run ClickHouse inside Docker follow the guide on [Docker Hub](https://hub.docker.com/r/clickhouse/clickhouse-server/). Those images use official `deb` packages inside.
You can install ClickHouse on Linux using a single portable binary from the latest commit of the `master` branch: [https://builds.clickhouse.com/master/amd64/clickhouse].
For non-Linux operating systems and for AArch64 CPU architecture, ClickHouse builds are provided as a cross-compiled binary from the latest commit of the `master` branch (with a few hours delay).
Run `sudo ./clickhouse install` to install ClickHouse system-wide (also with needed configuration files, configuring users etc.). Then run `sudo clickhouse start` commands to start the clickhouse-server and `clickhouse-client` to connect to it.
You can compile packages and install them or use programs without installing packages. Also by building manually you can disable SSE 4.2 requirement or build for AArch64 CPUs.
You’ll need to create a data and metadata folders and `chown` them for the desired user. Their paths can be changed in server config (src/programs/server/config.xml), by default they are:
By default, it connects to `localhost:9000` on behalf of the user `default` without a password. It can also be used to connect to a remote server using `--host` argument.
