[translate to english] Build a CubeFS cluster on a single machine

[heymingwei]

I need someone to translate this article to english, thanks

https://mp.weixin.qq.com/s/_-CcKBbNBRyR6mHkfNZIdw

[hooklee2000]

with OpenAI ( register only by email and free to use) ChatGPT 3.5, you can say "翻译成英语" then past the chinese content. I think the result can be acceptable.

CubeFS Quick Start in Ten Minutes | Single-node Deployment of CubeFS Cluster Original by Xuewei Zeng, CubeFS, March 7, 2023, 20:00, Guangdong Author Introduction

Xuewei Zeng, one of the CubeFS Maintainers, responsible for the development of subsystems such as multi-replication and metadata management, currently working at OPPO.

For new learners who want to quickly set up a CubeFS cluster to get started, the community currently provides two solutions:

Distributed deployment, preparing at least 3 or more machines to build the cluster Single-node deployment, setting up a single-node docker cluster using Docker Both approaches have a certain level of complexity, so to further reduce the learning curve, this article introduces how to quickly set up a single-node CubeFS cluster without using Docker.

Note: This article describes the process using a 64-bit CentOS 7.6 system as an example, with the following manual deployment tutorial as reference.

Two methods are provided here to obtain binaries, direct download or manual compilation.

01 Binary Download

bash Copy code git clone https://github.com/cubefs/cubefs.git cd ./cubefs wget https://ocs-cn-north1.heytapcs.com/cubefs/bin/linux_centos_7_6_amd64.tar.gz tar linux_centos_7_6_amd64.tar.gz

tree ./build/bin build/bin/ ├── cfs-cli ├── cfs-client └── cfs-server Note: Use ./build/bin/cfs-server -v to check. If there is an error, you will need to compile the binaries manually.

02 Manual Compilation

Refer to the CubeFS manual deployment document compilation method

Prepare the compilation environment Dependencies Version Requirement gcc-c++ 4.8.5 and above CMake 3.1 and above Go 1.16 and above bzip2-devel 1.0.6 and above First, download the latest module code from GitHub, for example, download it to a local directory

bash Copy code git clone https://github.com/cubefs/cubefs.git Enter the cubefs directory and execute compilation cd ./cubefs make server make client make cli Compiled binaries are located in the build/bin directory The deployment script creates four new IPs using network card aliases for service deployment

Deployment service topology diagram as follows

Execute the deployment script to start the cluster Require root permissions Able to use ifconfig Machine memory is above 4G The remaining space of the disk corresponding to /home/data is preferably above 20G

bash Copy code sh ./shell/depoly.sh /home/data bond0 /home/data: A directory on the local machine used to save cluster running logs and data, as well as default configuration files for each module bond0: The name of the local machine's network card, as follows, needs to be filled in according to the local situation

View cluster status Manage the cubefs cluster using the cfs-cli tool

bash Copy code ./build/bin/cfs-cli cluster info [Cluster] Cluster name : cfs_dev Master leader : 172.16.1.101:17010 Auto allocate : Enabled MetaNode count : 4 MetaNode used : 0 GB MetaNode total : 21 GB DataNode count : 4 DataNode used : 191 GB DataNode total : 369 GB Volume count : 2 EbsAddr : loadFactor : 0 BatchCount : 0 MarkDeleteRate : 0 DeleteWorkerSleepMs: 0 AutoRepairRate : 0 Note: Due to multiple data&meta nodes deployed on the same machine, the statistics of meta&data used and total are biased.

Create volume

bash Copy code ./build/bin/cfs-cli volume create ltptest ltp

View volume information

./build/bin/cfs-cli volume info ltptest Start the client /home/data is the directory passed in when starting above

bash Copy code ./build/bin/cfs-client -c /home/data/conf/client.conf df -h to check if the mount is successful Unmount

bash Copy code umount /home/cfs/client/mnt

Or force unmount

umount -l /home/cfs/client/mnt Stop the service This will stop the server and cubefs mounts

bash Copy code sh ./shell/stop.sh The above represents the successful establishment of the entire cluster and normal use.

The above installation tutorial is suitable for students who want to get a preliminary understanding of CubeFS and for students who want to further test performance, it is recommended to use the distributed deployment solution.

[heymingwei]

please make a PR, thanks @hooklee2000

[hooklee2000]

which directory startSeries , technicalInsights or bestPractices?