通用参数配置
在 matrixone/etc/launch-tae-CN-tae-DN/ 目录有三个配置文件 cn.toml、dn.toml 和 log.toml。
各个配置文件中所含参数解释如下:
cn.toml
参数 | 参数解释 |
---|---|
service-type = “CN” | 节点类型 |
data-dir = “./mo-data” | 默认数据目录 |
[log] | |
level = “info” | 日志级别,可修改为 info/debug/error/faltal |
format = “console” | 日志格式 |
max-size = 512 | 日志默认长度 |
[hakeeper-client] | HAkeeper 默认地址与端口,不建议更改 |
service-addresses = [ “127.0.0.1:32001”, ] | |
[[fileservice]] | fileservice 配置,不建议更改 |
name = “LOCAL” | fileservice 存储类型,本地存储 |
backend = “DISK” | fileservice 后端介质,磁盘 |
[[fileservice]] | |
name = “S3” | fileservice 存储类型,S3 |
backend = “DISK” | fileservice 后端介质,磁盘 |
data-dir = “mo-data/s3” | s3 存储数据路径 |
[[fileservice]] | |
name = “ETL” | fileservice 存储类型,ETL |
backend = “DISK-ETL” | fileservice 后端介质,DISK-ETL |
[observability] | 可观测性参数,默认不开启 |
disableTrace = true | |
disableMetric = true | |
[cn] | cn 节点的编号,不可修改 |
uuid = “dd1dccb4-4d3c-41f8-b482-5251dc7a41bf” | |
[cn.Engine] | cn 节点的存储引擎,分布式 tae,不可修改 |
type = “distributed-tae” |
dn.toml
参数 | 参数解释 |
---|---|
service-type = “DN” | 节点类型 |
data-dir = “./mo-data” | 默认数据目录 |
[log] | |
level = “info” | 日志级别,可修改为info/debug/error/faltal |
format = “console” | 日志格式 |
max-size = 512 | 日志默认长度 |
[hakeeper-client] | HAkeeper 默认地址与端口,不建议更改 |
service-addresses = [ “127.0.0.1:32001”, ] | |
[[fileservice]] | fileservice 配置,不建议更改 |
name = “LOCAL” | fileservice 存储类型,本地存储 |
backend = “DISK” | fileservice 后端介质,磁盘 |
[[fileservice]] | |
name = “S3” | fileservice 存储类型,S3 |
backend = “DISK” | fileservice 后端介质,磁盘 |
data-dir = “mo-data/s3” | s3 存储数据路径 |
[[fileservice]] | |
name = “ETL” | fileservice 存储类型,ETL |
backend = “DISK-ETL” | fileservice 后端介质,DISK-ETL |
[dn] | |
uuid = “dd4dccb4-4d3c-41f8-b482-5251dc7a41bf” | dn 的 uuid,不可修改 |
[dn.Txn.Storage] | dn 事务后端的存储引擎,不可修改 |
backend = “TAE” | |
log-backend = “logservice” | |
[dn.Ckp] | dn 的 checkpoint 相关参数,不建议更改 |
flush-interval = “60s” | 内部刷新间隔 |
min-count = 100 | checkpoint 最小个数 |
scan-interval = “5s” | 内部扫描间隔 |
incremental-interval = “180s” | checkpoint 自增间隔 |
global-interval = “100000s” | 全局 checkpoint 间隔 |
log.toml
参数 | 参数解释 |
---|---|
service-type = “LOG” | 节点类型 |
data-dir = “./mo-data” | 默认数据目录 |
[log] | |
level = “info” | 日志级别,可修改为info/debug/error/faltal |
format = “console” | 日志格式 |
max-size = 512 | 日志默认长度 |
[[fileservice]] | |
name = “S3” | fileservice 存储类型,S3 |
backend = “DISK” | fileservice 后端介质,磁盘 |
data-dir = “mo-data/s3” | s3 存储数据路径 |
[[fileservice]] | |
name = “ETL” | fileservice 存储类型,ETL |
backend = “DISK-ETL” | fileservice 后端介质,DISK-ETL |
[observability] | 监控相关参数 |
statusPort = 7001 | 预留普罗米修斯的监控端口 |
enableTraceDebug = false | 开启 trace 功能的 dbug 模式 |
[hakeeper-client] | HAkeeper 默认地址与端口,不建议更改 |
service-addresses = [ “127.0.0.1:32001”, ] | |
[logservice] | logservice 的相关参数,不可修改 |
deployment-id = 1 | logservice 的部署 id |
uuid = “7c4dccb4-4d3c-41f8-b482-5251dc7a41bf” | logservice 的节点 id |
raft-address = “127.0.0.1:32000” | raft 协议使用的地址 |
logservice-address = “127.0.0.1:32001” | logservice 服务地址 |
gossip-address = “127.0.0.1:32002” | gossip 协议的地址 |
gossip-seed-addresses = [ “127.0.0.1:32002”, ] | gossip 协议的种子节点地址 |
gossip-allow-self-as-seed = true | 是否允许 gossip 协议用本节点做种子节点 |
[logservice.BootstrapConfig] | bootstrap 相关参数,不可修改 |
bootstrap-cluster = true | bootstrap 是否集群启动 |
num-of-log-shards = 1 | logservice 的分片数 |
num-of-dn-shards = 1 | dn 的分片数 |
num-of-log-shard-replicas = 1 | logservice 分片的副本数 |
init-hakeeper-members = [ “131072:7c4dccb4-4d3c-41f8-b482-5251dc7a41bf”, ] | 初始化 hakeeper 的成员 |