greptimedb/config/datanode.example.toml

# Node running mode, see `standalone.example.toml`.
mode = "distributed"
# The datanode identifier, should be unique.
node_id = 42
# gRPC server address, "127.0.0.1:3001" by default.
rpc_addr = "127.0.0.1:3001"
# Hostname of this node.
rpc_hostname = "127.0.0.1"
# The number of gRPC server worker threads, 8 by default.
rpc_runtime_size = 8
# Start services after regions have obtained leases.
# It will block the datanode start if it can't receive leases in the heartbeat from metasrv.
require_lease_before_startup = false

# Initialize all regions in the background during the startup.
# By default, it provides services after all regions have been initialized.
initialize_region_in_background = false

[heartbeat]
# Interval for sending heartbeat messages to the Metasrv, 3 seconds by default.
interval = "3s"

# Metasrv client options.
[meta_client]
# Metasrv address list.
metasrv_addrs = ["127.0.0.1:3002"]
# Heartbeat timeout, 500 milliseconds by default.
heartbeat_timeout = "500ms"
# Operation timeout, 3 seconds by default.
timeout = "3s"
# Connect server timeout, 1 second by default.
connect_timeout = "1s"
# `TCP_NODELAY` option for accepted connections, true by default.
tcp_nodelay = true

# WAL options.
# Currently, users are expected to choose the wal through the provider field. 
# When a wal provider is chose, the user should comment out all other wal config 
# except those corresponding to the chosen one.
[wal]
# WAL data directory
provider = "raft_engine"

# Raft-engine wal options, see `standalone.example.toml`.
# dir = "/tmp/greptimedb/wal"
file_size = "256MB"
purge_threshold = "4GB"
purge_interval = "10m"
read_batch_size = 128
sync_write = false

# Kafka wal options, see `standalone.example.toml`.
# broker_endpoints = ["127.0.0.1:9092"]
# max_batch_size = "4MB"
# linger = "200ms"
# produce_record_timeout = "100ms"
# backoff_init = "500ms"
# backoff_max = "10s"
# backoff_base = 2
# backoff_deadline = "5mins"

# Storage options, see `standalone.example.toml`.
[storage]
# The working home directory.
data_home = "/tmp/greptimedb/"
# Storage type.
type = "File"
# TTL for all tables. Disabled by default.
# global_ttl = "7d"

# Cache configuration for object storage such as 'S3' etc.
# The local file cache directory
# cache_path = "/path/local_cache"
# The local file cache capacity in bytes.
# cache_capacity = "256MB"

# Custom storage options
#[[storage.providers]]
#type = "S3"
#[[storage.providers]]
#type = "Gcs"

# Mito engine options
[[region_engine]]
[region_engine.mito]
# Number of region workers
num_workers = 8
# Request channel size of each worker
worker_channel_size = 128
# Max batch size for a worker to handle requests
worker_request_batch_size = 64
# Number of meta action updated to trigger a new checkpoint for the manifest
manifest_checkpoint_distance = 10
# Whether to compress manifest and checkpoint file by gzip (default false).
compress_manifest = false
# Max number of running background jobs
max_background_jobs = 4
# Interval to auto flush a region if it has not flushed yet.
auto_flush_interval = "1h"
# Global write buffer size for all regions.
global_write_buffer_size = "1GB"
# Global write buffer size threshold to reject write requests (default 2G).
global_write_buffer_reject_size = "2GB"
# Cache size for SST metadata (default 128MB). Setting it to 0 to disable the cache.
sst_meta_cache_size = "128MB"
# Cache size for vectors and arrow arrays (default 512MB). Setting it to 0 to disable the cache.
vector_cache_size = "512MB"
# Cache size for pages of SST row groups (default 512MB). Setting it to 0 to disable the cache.
page_cache_size = "512MB"
# Buffer size for SST writing.
sst_write_buffer_size = "8MB"
# Parallelism to scan a region (default: 1/4 of cpu cores).
# - 0: using the default value (1/4 of cpu cores).
# - 1: scan in current thread.
# - n: scan in parallelism n.
scan_parallelism = 0
# Capacity of the channel to send data from parallel scan tasks to the main task (default 32).
parallel_scan_channel_size = 32

# Log options, see `standalone.example.toml`
# [logging]
# dir = "/tmp/greptimedb/logs"
# level = "info"

# Datanode export the metrics generated by itself
# encoded to Prometheus remote-write format
# and send to Prometheus remote-write compatible receiver (e.g. send to `greptimedb` itself)
# This is only used for `greptimedb` to export its own metrics internally. It's different from prometheus scrape.
# [export_metrics]
# whether enable export metrics, default is false
# enable = false
# The url of metrics export endpoint, default is `frontend` default HTTP endpoint.
# endpoint = "127.0.0.1:4000"
# The database name of exported metrics stores, user needs to specify a valid database
# db = ""
# The interval of export metrics
# write_interval = "30s"
# HTTP headers of Prometheus remote-write carry
# headers = {}