Boot Parameters for standalone installation

There are three configuration files cn.toml, dn.toml and log.toml in the matrixone/etc/launch-tae-CN-tae-DN/ directory.

The parameters contained in each configuration file are explained as follows:

cn.toml

ParametersDescription
service-type = “CN”Node Type
data-dir = “./mo-data”Default data directory
[log]
level = “info”Log level, can be modified to info/debug/error/faltal
format = “console”Log format
max-size = 512Log default length
[hakeeper-client]HAkeeper default address and port, not recommended to change
service-addresses = [
“127.0.0.1:32001”,
]
[metacache]Metadata Cache
memory-capacity = “512MB”Set the cache size for metadata, default is 512MB. The data caching feature results in slower first-time queries, but subsequent queries will be faster.
[[fileservice]]fileservice, not recommended to change
name = “LOCAL”fileservice storage type, local storage
backend = “DISK”fileservice backend, disk
[[fileservice]]
name = “SHARED”fileservice storage type, S3
backend = “DISK”fileservice backend, disk
data-dir = “mo-data/s3”s3 storage data directory
[fileservice.cache]
memory-capacity = “512MB”cache memory size used by fileservice
disk-capacity = “8GB”cache disk size used by fileservice
disk-path = “mo-data/file-service-cache”fileservice disk cache path
disk-min-evict-interval = “7m”Interval for disk cache recovery, in seconds (s) or minutes (m)
disk-evict-target = 0.8Target capacity for disk cache recovery, parameter is the ratio of template capacity to total capacity
[[fileservice]]
name = “ETL”fileservice storage type, ETL
backend = “DISK-ETL”fileservice backend, DISK-ETL
[observability]Observability parameter, disabled by default
disableTrace = true
disableMetric = true
[cn]cn code number, cannot be modified
uuid = “dd1dccb4-4d3c-41f8-b482-5251dc7a41bf”
[cn.Engine]Storage engine of the cn node, distributed tae, cannot be modified
type = “distributed-tae”
[cn.txn]
enable-sacrificing-freshness = falseWhen set to true and push mode, this parameter does not guarantee that a transaction can see the latest committed data. Instead, it uses the latest Logtail commit timestamp received by the current CN as the transaction start time. This setting ensures that a transaction on the same database connection can see the data written by its previously committed transactions.
enable-cn-based-consistency = falseWhen the above parameter is set to true, it ensures external consistency on the same CN. When a transaction starts, it can see the data written by previously committed transactions.

dn.toml

ParametersDescription
service-type = “DN”Node Type
data-dir = “./mo-data”Default data directory
[log]
level = “info”Log level, can be modified to info/debug/error/faltal
format = “console”Log format
max-size = 512Log default length
[hakeeper-client]HAkeeper default address and port, not recommended to change
service-addresses = [
“127.0.0.1:32001”,
]
[metacache]Metadata Cache
memory-capacity = “512MB”Set the cache size for metadata, default is 512MB. The data caching feature results in slower first-time queries, but subsequent queries will be faster.
[[fileservice]]fileservice, not recommended to change
name = “LOCAL”fileservice storage type, local storage
backend = “DISK”fileservice backend, disk
[[fileservice]]
name = “SHARED”fileservice storage type, S3
backend = “DISK”fileservice backend, disk
data-dir = “mo-data/s3”s3 storage data directory
[fileservice.cache]
memory-capacity = “512MB”cache memory size used by fileservice
disk-capacity = “8GB”cache disk size used by fileservice
disk-path = “mo-data/file-service-cache”fileservice disk cache path
disk-min-evict-interval = “7m”Interval for disk cache recovery, in seconds (s) or minutes (m)
disk-evict-target = 0.8Target capacity for disk cache recovery, parameter is the ratio of template capacity to total capacity
[[fileservice]]
name = “ETL”fileservice storage type, ETL
backend = “DISK-ETL”fileservice backend, DISK-ETL
[dn]
uuid = “dd4dccb4-4d3c-41f8-b482-5251dc7a41bf”uuid of dn, cannot be modified
[dn.Txn.Storage]The storage engine of the dn transaction backend, cannot be modified
backend = “TAE”
log-backend = “logservice”
[dn.Txn]
incremental-dedup = falseIf set to false, DN initiates deduplication for all data. If set to true, DN will only deduplicate primary key data after the snapshot timestamp.
[dn.Ckp]the checkpoint related parameters of dn, not recommended to change
flush-interval = “60s”internal refresh interval
min-count = 100Minimum number of checkpoints
scan-interval = “5s”internal scan interval
incremental-interval = “180s”checkpoint increment interval
global-min-count = 60The global minimum number of dn checkpoints
[dn.LogtailServer]
listen-address = “0.0.0.0:32003”logtail listening port
service-address = “127.0.0.1:32003”logtail internal access address
rpc-max-message-size = “16KiB”maximum rpc message size used by logtail
rpc-payload-copy-buffer-size = “16KiB”rpc copy buffer size
rpc-enable-checksum = truewhether to enable rpc checksum
logtail-collect-interval = “2ms”logtail statistics collection interval
logtail-response-send-timeout = “10s”logtail sending timeout
max-logtail-fetch-failure = 5The maximum number of failures allowed by fetching logtail

log.toml

ParametersDescription
service-type = “LOG”Node Type
data-dir = “./mo-data”Default data directory
[log]
level = “info”Log level, can be modified to info/debug/error/faltal
format = “console”Log format
max-size = 512Log default length
[[fileservice]]fileservice configuration, not recommended to change
name = “LOCAL”fileservice storage type, local storage
backend = “DISK”fileservice backend media, disk
[[fileservice]]
name = “SHARED”fileservice storage type, S3
backend = “DISK”fileservice backend, disk
data-dir = “mo-data/s3”s3 storage data directory
[fileservice.cache]
memory-capacity = “512MB”cache memory size used by fileservice
disk-capacity = “8GB”cache disk size used by fileservice
disk-path = “mo-data/file-service-cache”fileservice disk cache path
disk-min-evict-interval = “7m”Interval for disk cache recovery, in seconds (s) or minutes (m)
disk-evict-target = 0.8Target capacity for disk cache recovery, parameter is the ratio of template capacity to total capacity
[[fileservice]]
name = “ETL”fileservice storage type, ETL
backend = “DISK-ETL”fileservice backend, DISK-ETL
[observability]Monitor parameters
statusPort = 7001Reserve the monitoring port of Prometheus
enableTraceDebug = falseEnable the dbug mode of the trace
[hakeeper-client]HAkeeper default address and port, not recommended to change
service-addresses = [
“127.0.0.1:32001”,
]
[logservice]The parameters of logservice, cannot be modified
deployment-id = 1Deployment ID of logservice
uuid = “7c4dccb4-4d3c-41f8-b482-5251dc7a41bf”Node ID of logservice
raft-address = “127.0.0.1:32000”The address of the raft protocol
logservice-address = “127.0.0.1:32001”logservice address
gossip-address = “127.0.0.1:32002”The address of the gossip protocol
gossip-seed-addresses = [
“127.0.0.1:32002”,
]
The root node address of the gossip protocol
gossip-allow-self-as-seed = trueWhether to allow the gossip protocol to use this node as a root node
[logservice.BootstrapConfig]Bootstrap parameters, cannot be modified
bootstrap-cluster = trueWhether bootstrap cluster launchs
num-of-log-shards = 1The number of shards of logservice
num-of-dn-shards = 1The number of shards of dn
num-of-log-shard-replicas = 1The number of replicas of the logservice shard
init-hakeeper-members = [
“131072:7c4dccb4-4d3c-41f8-b482-5251dc7a41bf”,
]
Initialize members of hakeeper