rqlite Design High-level design Design presentations Blog posts Other design details Raft SQLite Log Compaction and Truncation rqlite Design Learn about the design and im...

副本层 总体概览 功能组件 RAFT SnapShot Lease Holder Lease-读场景 Lease-写场景 副本层 总体概览 一个存储设备(store)可以 存储若干Range，被一个 Rocksdb实例管理 一个range是一段有序键值 区间【K1，K2】的数据分片， 是DRDB存储管理的最小单位 每个...

Geo-Distributed Deployment Topology Topology information Topology templates Key parameters TiKV parameters PD parameters Geo-Distributed Deployment Topology This document t...

Consensus Protocol Raft Protocol Overview Raft in Consul Consistency Modes Deployment Table Consensus Protocol Consul uses a consensus protocol ) to provide Consistency (as ...

TiDB Best Practices on Public Cloud Use a dedicated disk for Raft Engine Choose appropriate disks for Raft Engine on public clouds Middle-range disk High-end disk Example 1: Run...

对于 v1.0 前版本，OVN 的 RAFT 模式存在机器重启后无法重新恢复的问题，建议升级至 1.0 后的版本。在 1.0 前版本进行修复需要重启启动所有容器网络下的 Pod，并且重启机器仍然有概率出现该问题。 检查对应机器的 /var/log/openvswitch/ovsdb-server-nb.log 和 /var/log/openvswitch...

Replication with Raft consensus Strong write consistency Tunable read consistency Read replicas Replication with Raft consensus DocDB replicates data in order to survive fai...

Node Placement Driver (PD) Peer Raft Region Store Transaction This document explains some important terminologies of TiKV and PD. Node A TiKV Node is a physical node in ...

海量 Region 集群调优最佳实践 Raftstore 的工作流程 性能问题 性能监控 性能优化方法 方法一：增加 TiKV 实例 方法二：调整 raft-base-tick-interval 方法三：提高 Raftstore 并发数 方法四：开启 Hibernate Region 功能 方法五：开启 Region Merge 方法六：...

Performance Performance Factors Disk Network Improving Performance Batching Queued Writes Use more powerful hardware Use a memory-backed filesystem Placing the SQLite databa...