About "pending compaction bytes" and "write stall"
Open
component/tikv
type/question
- Dominant language
- Go
- Stars
- 40.5k
- Forks
- 6.2k
- PR merge metrics
- PR metrics pending
Description
File: [/release-4.0/tidb-troubleshooting-map.md](https://docs.pingcap.com/zh/tidb/v4.0/tidb-troubleshooting-map)
如果 pending compaction bytes 达到该阈值，**RocksDB 会放慢写入速度。**默认值 64GB。
1、这个”RocksDB会放慢写入速度“，具体是怎么个放慢的方法？ 是pending compaction bytes达到了64GB，会有另外一个参数去控制写入的流量吗？
2、所谓的write stall，是pending compaction bytes 达到64GB，就属于write stall了吗？
Contributor guide
Assessment
This issue has not been assessed yet.