tencent cloud

Getting Started
Download PDF
Last updated: 2025-07-30 16:56:45
Getting Started
Last updated: 2025-07-30 16:56:45
Download PDF

Creating Cluster

1. Log in to the TCHouse-C overview page and click Buy Now or log in to the TCHouse-C console and click New Cluster with your Tencent Cloud account.
2. On the purchase page, configure and purchase a cluster as prompted. For more information on configuration items, see Configuration Items.


Configuration items

Configuration Item
Description
Billing Mode
Pay-as-you-go: postpaid, where a bill is generated hourly based on resource usage and then you pay for what you use.
Monthly subscription: Prepaid Billing, You need to pay the fees when creating an instance.
Region
Currently, TCHouse-C is available in the Beijing, Shanghai, Guangzhou, Chengdu, Nanjing, Hong Kong, as well as Singapore, Jakarta, Tokyo, and Frankfurt regions. We recommend you select a region closest to your users, and you cannot change the region after the purchase.
Availability Zone
Select availability zones in different regions as needed on the purchase page.
Network
A VPC is an isolated, highly secure, and dedicated network environment. You can create a VPC and subnet or select an existing one.
High Availability
In HA mode, each shard has two replicas; in non-HA mode, each shard has only one replica, where the entire cluster will fail if the replica fails. Therefore, we recommend you use the HA mode for production environments.
Compute Node Type
There are three types of compute nodes, the actual resource reserves vary by region, and the specifics are subject to the information displayed on the purchase page:
Standard: 4-core 16 GB, 8-core 32 GB, 16-core 64 GB, 32-core 128 GB, 64-core 256 GB, 90-core 224 GB, and 128-core 256 GB.
Storage-Optimized: 32-core 128 GB (with twelve 3720 GB SATA HDDs) and 64-core 256 GB (with twenty-four 3720 GB SATA HDDs), 84-core 320 GB (with twenty-four 3720 GB SATA HDDs).
High-Performance: 32-core 128 GB (with two 3570 GB NVMe SSDs), 64-core 256 GB (with four 3570 GB NVMe SSDs), and 84-core 320 GB (with four 3570 GB NVMe SSDs).
The higher the specification, the better the performance. You can select an appropriate specification as needed.
ZooKeeper Node Type
There are 4-core 16 GB, 8-core 32 GB, 16-core 64 GB, 24-core 96 GB, 32-core 128 GB, 64-core 256 GB, 90-core 224 GB, and 128-core 256 GB ZooKeeper nodes. The heavier the load, the higher the specification needed. You can select an appropriate specification as needed.
Note:
You can enable dedicated Grafana monitoring, cluster logging, tiered storage of hot/cold data, and auto-renewal features as needed.

Viewing Cluster Information

After the cluster is created, go to the TCHouse-C console, select the region where the cluster resides, and view the cluster status and information as shown below:




Using ClickHouse

Import a data file to a ClickHouse cluster and view the imported data. Prepare the following account.csv file:
AccountId, Name, Address, Year
1, 'GHua', 'WuHan Hubei', 1990
2, 'SLiu', 'ShenZhen Guangzhou', 1991
3, 'JPong', 'Chengdu Sichuan', 1992

Connecting to cluster

Download a ClickHouse Client, recommended download clickhouse-client-23.3.8.21.x86_64.rpm and clickhouse-common-static-23.3.8.21.x86_64.rpm (for Centos) .
Install the client.
rpm -ivh *.rpm
Access the cluster from the client.
View the node IP address in the console and select the TCP port 9000.
clickhouse-client -hxxx.xxx.xxx.xxx --port 9000



Select the HTTP port 8123 and get the specific access IP address in Cluster Access Address (HTTP) on the cluster details page.
Query and confirm the engine version of the cluster.
echo "select version()=21.3.9.83"  |  curl  'http://xxx.xxx.xxx.xxx:8123/' --data-binary @-
echo "select version()"  |  curl  'http://xxx.xxx.xxx.xxx:8123/' --data-binary @-



Query the system cluster.
echo "select * from system.clusters"  |  curl  'http://xxx.xxx.xxx.xxx:8123/' --data-binary @-




Creating data table

Use the ClickHouse client to connect to the cluster and create databases and tables.
Create a database in HA mode
CREATE DATABASE IF NOT EXISTS testdb ON CLUSTER default_cluster;



Create a table in HA mode
CREATE TABLE testdb.account ON CLUSTER default_cluster(accountid UInt16,name String,address String,year UInt64) ENGINE =ReplicatedMergeTree('/clickhouse/tables/{layer}-{shard}/testdb/account', '{replica}') ORDER BY (accountid);



Create a database in non-HA mode
CREATE DATABASE IF NOT EXISTS testdb ON CLUSTER default_cluster;



Create a table in non-HA mode
CREATE TABLE testdb.account ON CLUSTER default_cluster(accountid UInt16, name String, address String, year UInt64) ENGINE =MergeTree() ORDER BY (accountid);




Importing data

Place the prepared data in the /data directory of the CVM instance connected to the ClickHouse cluster and run the following command to import the data.
cat /data/account.csv | clickhouse-client - hxxx.xxx.xxx.xxx --database=testdb --query="INSERT INTO account FORMAT CSVWithNames"

Querying data

select * from testdb.account;



Was this page helpful?
You can also Contact Sales or Submit a Ticket for help.
Yes
No

Feedback