MinIO
2022-03-03 10:03:30 13 举报AI智能生成
登录查看完整内容
MinIO 对象存储
作者其他创作
大纲/内容
Efficient S3 compatibility
Erasure Coding 纠删码
数据治愈
BitRot Protection
Identity and Access Management
权限管理 IAM
Encryption
Data Lifecycle & Tiering
Immutability(Locking)
Versioning
Notification
Scalability
Replication
MinIO has no separate metadata store
Metadata
特性
5 TiB
Maximum object size
63
Maximum length for bucket names
1024
Maximum length for object names
255
Maximum length for '/' separated object name segment
Limits
更换、重新mount磁盘即可
磁盘损坏
重新启一个节点即可
node 故障
Recover
用于构建高性能对象存储服务 的 软件系统
兼容 AWS S3
是什么？
高性能
可扩展
简约
设计目标
MinIO Server
MinIO Client
Cluster Installation & Management
MinIO Operator
GUI Dashboard
MinIO Console
Java/Go/Python/Js
MinIO SDK
组件
单节点单磁盘
快速体验
nfs
Gateway
4 to 16 drives per set. (单节点多磁盘也可以启用纠删码)
Each object is written to a single EC set
18 drives: 9 * 2
24 drives: 12 * 2
Minio uses the largest possible EC set size
The drives should all be of approximately the same size.
N 最大取 ES中磁盘数的一半 (即 最小利用率为 50%)
Erasure Code Parity (EC:N)
ES(erasure-coding sets)= Erasure Code Stripe Size
Each node shoud have the same number of drives
NTP: 节点间时间差应小于 15min
minio server http://host{1...n}/export{1...m}
N nodes & M drives/node
Nodes
pool 1: host{1...n}
pool 2: host{o...z}
minio server http://host{1...n}/export{1...m} http://host{o...z}/export{1...m}
New object upload requests automatically start using the least used cluster.
Each pool should have the same erasure coding parity configuration
Cluster Expanding
Direct-Attached Storage (DAS) with XFS for best performance.
drives
Distributed(drives >= 4)
部署模式
MinIO
1. 确保前后软硬件环境基本一致2. mc admin heal/monitor 观察minio自我修复
0 条评论
回复 删除
下一页
