TiDB Passes Jepsen Test for Snapshot Isolation and Single-Key Linearizability
TiDB's first official Jepsen Test report is published. This post introduces some additional context to the test results and PingCAP's thoughts on what's next.
Design and Implementation of Golang Failpoints
This post introduces design and implementation of Golang Failpoint.
Try to Fix Two Linux Kernel Bugs While Testing TiDB Operator in K8s
This post introduces how the EE team at PingCAP tackled two Linux kernel bugs while testing TiDB Operator in K8s. PingCAP engineers also hope the K8s community, RHEL and CentOS can help fix these bugs thoroughly in the near future.
How TiKV Uses “Lease Read” to Guarantee High Performances, Strong Consistency and Linearizability
This post discusses Raft Log Read, `ReadIndex` Read, and Lease Read, and why TiKV adopts the Lease Read approach.
9 Why’s to Ask When Evaluating a Distributed Database
Ed Huang, the CTO of PingCAP, has summarized into "9 why's" the most frequently asked questions by engineers when they look at a distributed database; he also gives his answers to these questions in this post to help make decision-making of engineers a bit easier.
Implement Raft in Rust
As an open-source distributed scalable HTAP database, TiDB uses the Raft Consensus Algorithm in its distributed transactional key-value storage engine, TiKV, to ensure data consistency, auto-failover, and fault tolerance. TiDB has thus far been used by more than 200 companies in their production environments in a wide range of industries, from e-commerce and food delivery, to fintech, media, gaming, and travel.
From Chaos to Order — Tools and Techniques for Testing TiDB, A Distributed NewSQL Database
As an open source distributed NewSQL Hybrid Transactional/Analytical Processing (HTAP) database, TiDB contains the most important asset of our customers--their data. One of the fundamental and foremost requirements of our system is to be fault-tolerant. But how do you ensure fault tolerance in a distributed database? This article covers the top fault injection tools and techniques in Chaos Engineering, as well as how to execute Chaos practices in TiDB.
How to Do Performance Tuning on TiDB, a Distributed NewSQL Database
Doing performance tuning on distributed systems is no joking matter. It's much more complicated than on a single node server, and bottlenecks can pop up anywhere, from system resources in a single node or subcomponent, to cooperation between nodes, to even network bandwidth. Performance tuning is a practice that aims to find these bottlenecks and address them, in order to reveal more bottlenecks and address them as well, until the system reaches an optimal performance level. In this article, I will share some best practices on how to tune "write" operations in TiDB to achieve maximum performance.
Tick or Tock? Keeping Time and Order in Distributed Databases
At re:Invent 2017, Amazon Web Services (AWS) announced Amazon Time Sync Service which is a highly accurate and reliable time reference that is natively accessible from Amazon EC2 instances. It is much like the Google TrueTime which was published in 2012. Why do Google and AWS both want to make efforts to provide global time service? Is there any inspiration for building distributed database? This topic is important to think about.
A TiKV Source Code Walkthrough – Raft Optimization
Paxos or Raft is frequently used to ensure data consistency in the distributed computing area. But Paxos is known for its complexity and is rather difficult to understand while Raft is very simple. Therefore, a lot of emerging databases tend to use Raft as the consensus algorithm at its bottom layer. TiKV is no exception.
Why did we choose Rust over Golang or C/C++ to develop TiKV?
Every developer has his/her favorite programming language. For the TiKV team members, it's Rust.
RocksDB in TiKV
This is the speech Siddon Tang gave at the RocksDB meetup on August 28, 2017.