These docs are for Cribl Stream 4.15 and are no longer actively maintained.
See the latest version (4.16).
service.yml
service.yml maintains configuration for Cribl Stream service processes.
In the UI, you can configure some of these settings on the Leader at Settings > Global > System > Service Processes > Services.
Be careful about reducing the predefined
memoryLimitvalues. Extremely low values can prevent some Cribl Stream components from functioning.
service.yml
# Service configuration - Configuration for various Cribl services
# Available service types: connections, metrics, notifications, stream_connections, edge_connections, jobs, outpost
connections:
# Connections listener number of processes - Negative values are offsets from the number of available cores
# on the host.
# [number; default: 1]
procs:
# Connections listener heap memory limit - Heap memory limit for the connection listener processes
# [string; default: 2GB]
memoryLimit:
metrics:
# Metrics number of processes - Single process; values above 1 will be treated as 1, negative values are
# offsets from the number of available cores on the host.
# [number; default: 1]
procs:
# Metrics heap memory limit - Heap memory limit for the metrics process
# [string; default: 2GB]
memoryLimit:
notifications:
# Notifications number of processes - Single process; values above 1 will be treated as 1, negative values
# are offsets from the number of available cores on the host.
# [number; default: 1]
procs:
# Notifications heap memory limit - Heap memory limit for the notifications process
# [string; default: 2GB]
memoryLimit:
stream_connections:
# Stream connection processes - Negative values are offsets from the number of available cores on the host.
# [number; default: 1]
procs:
# Auto scale - Do not modify this setting. Defines whether Leader can start the service once qualifying
# Worker node connects to it.
# [boolean]
autoScale:
# Stream connection memory limit - Heap memory limit for the Stream connection processes
# [string; default: 2GB]
memoryLimit:
edge_connections:
# Edge connection processes - Negative values are offsets from the number of available cores on the host.
# [number; default: 1]
procs:
# Edge auto scale - Do not modify this setting. Defines whether Leader can start the service once
# qualifying Edge node connects to it.
# [boolean]
autoScale:
# Edge connection memory limit - Heap memory limit for the Edge connection processes
# [string; default: 2GB]
memoryLimit:
# Connection proxy service - Proxy service that runs on a standby Leader node to forward traffic
# from port 4200 to the primary Leader.
connection_proxy:
# Connection proxy processes - Number of processes to spawn for connection proxy service.
# Set to 0 to disable the service. Negative values are offsets from the number of available cores on the host.
# [number; default: 0]
procs:
# Connection proxy receive buffer size - Socket receive buffer size for connection proxy processes in bytes.
# [number; default: 4096]
recvBufferSize:
# Connection proxy memory limit - Heap memory limit per process.
# [string; default: 2GB]
memoryLimit:
jobs:
# Jobs service processes - Number of processes to spawn for jobs service. Set to 0 to disable the service.
# [number; default: 1]
procs:
# Jobs service memory limit - Heap memory limit per process.
# [string; default: 2GB]
memoryLimit:
outpost:
# Outpost service processes - Number of processes to spawn for outpost service. Set to 0 to disable the
# service.
# [number; default: -2]
procs:
# Outpost service memory limit - Heap memory limit per process.
# [string; default: 2GB]
memoryLimit: