Subjects and Metrics

A rule watches one type of subject. This page lists the subject types, the properties you can filter and split on, and the most commonly used metrics. The rule editor always shows the complete list of metrics for the scope you choose, with a sensible default threshold for each.

Subject types

Environment subjects belong to one environment; rules on them have a Where it applies setting, and their alerts are grouped per environment. Workspace subjects belong to the workspace as a whole.

Subject	Properties to filter and split on	Belongs to
Handler	application name, handler type, component, message type, message name	Environment
Aggregate	application name, aggregate	Environment
Entity	application name, entity name	Environment
Processor	application name, processor name	Environment
Application	application name	Environment
Environment (Framework connections)	—	Environment
Axon Server	—	Environment
Axon Server node	node name	Environment
Axon Server disk	node name, disk path	Environment
Axon Server replication group	node name, replication group	Environment
Access token	token name; token ID for splitting	Workspace
Release	product, release version, release type, pre-release	Workspace
AI credits, Message credits	—	Workspace
AI credit bucket, Message credit bucket	bucket name; bucket ID for splitting	Workspace

Subject

Properties to filter and split on

Belongs to

Handler

application name, handler type, component, message type, message name

Environment

Aggregate

application name, aggregate

Environment

Entity

application name, entity name

Environment

Processor

application name, processor name

Environment

Application

application name

Environment

Environment (Framework connections)

—

Environment

Axon Server

—

Environment

Axon Server node

node name

Environment

Axon Server disk

node name, disk path

Environment

Axon Server replication group

node name, replication group

Environment

Access token

token name; token ID for splitting

Workspace

Release

product, release version, release type, pre-release

Workspace

AI credits, Message credits

—

Workspace

AI credit bucket, Message credit bucket

bucket name; bucket ID for splitting

Workspace

Access tokens and credit buckets are split by their ID rather than their name, so renaming one does not move its alert to a different inbox item.

Measures

Most metrics are compared as a value. Some offer a choice of how to compare them:

Percentile: Timing metrics, and other metrics measured per message, are distributions. Choose the minimum, the median, the ninetieth percentile, the ninety-fifth percentile, or the maximum.
Used and remaining: Metrics with a limit, such as heap or connections, can be compared as the amount used, the amount remaining, or either as a percentage of the limit.
On occurrence: Some events, such as a product release, have no threshold: the condition fires when the event happens.

Common metrics

Handlers, aggregates and entities

Metric	Description	Default threshold
Latency	Time to handle a message	> 200 ms
Error rate	Messages that failed, per minute	> 0
Throughput	Messages handled per minute	> 0
Lock time (aggregates)	Time spent acquiring the lock on the aggregate	> 100 ms
Load time (aggregates, entities)	Time to load the aggregate or entity from the event store	> 250 ms
Event commit time (aggregates, entities)	Time to commit the events the command applied	> 250 ms
Event stream size (aggregates, entities)	Number of events read from the event store to load the aggregate or entity	> 1000

Metric

Description

Default threshold

Latency

Time to handle a message

> 200 ms

Error rate

Messages that failed, per minute

> 0

Throughput

Messages handled per minute

> 0

Lock time (aggregates)

Time spent acquiring the lock on the aggregate

> 100 ms

Load time (aggregates, entities)

Time to load the aggregate or entity from the event store

> 250 ms

Event commit time (aggregates, entities)

Time to commit the events the command applied

> 250 ms

Event stream size (aggregates, entities)

Number of events read from the event store to load the aggregate or entity

> 1000

To alert when a handler stops receiving messages, use a throughput condition such as < 1. Other conditions on these subjects are only judged when messages are handled; see Only judge while handling in Warm up and cool down.

Processors

Metric	Description	Default threshold
Segment claim percentage	Share of the processor’s segments that are claimed by an instance	≠ 100%
Ingest latency	Time between an event being published and it being available for processing	> 1 minute
Commit latency	Time between an event being processed and the processor committing that	> 1 minute
Message processing latency	Time between the event being processed now and its publication: how far the processor is behind	> 1 minute
Dead-letter queue size	Number of messages in the processor’s dead-letter queue	> 0

Metric

Description

Default threshold

Segment claim percentage

Share of the processor’s segments that are claimed by an instance

≠ 100%

Ingest latency

Time between an event being published and it being available for processing

> 1 minute

Commit latency

Time between an event being processed and the processor committing that

> 1 minute

Message processing latency

Time between the event being processed now and its publication: how far the processor is behind

> 1 minute

Dead-letter queue size

Number of messages in the processor’s dead-letter queue

> 0

Applications

Metric	Description	Default threshold
Application replicas	Number of connected instances	< 1
Process CPU usage, Host CPU usage	CPU used by the application, or by the machine it runs on	> 80%
Heap	Heap memory used	> 80% used
CommandBus capacity, QueryBus capacity	Share of the bus capacity in use	> 80% used

Metric

Description

Default threshold

Application replicas

Number of connected instances

< 1

Process CPU usage, Host CPU usage

CPU used by the application, or by the machine it runs on

> 80%

Heap

Heap memory used

> 80% used

CommandBus capacity, QueryBus capacity

Share of the bus capacity in use

> 80% used

Axon Server

Metric	Description	Default threshold
Cluster connections	Client connections, against the cluster’s limit	≥ 80% used
Node disk space free	Free space on a node’s disk	< 1024 MB
Unhealthy cluster connections	Connections between nodes that are unhealthy	≠ 0
Healthy replicas, failing replicas	Replica health of a replication group	< 1 healthy, ≠ 0 failing
Un-applied entries	Entries a replication group has not applied yet	≥ 100

Metric

Description

Default threshold

Cluster connections

Client connections, against the cluster’s limit

≥ 80% used

Node disk space free

Free space on a node’s disk

< 1024 MB

Unhealthy cluster connections

Connections between nodes that are unhealthy

≠ 0

Healthy replicas, failing replicas

Replica health of a replication group

< 1 healthy, ≠ 0 failing

Un-applied entries

Entries a replication group has not applied yet

≥ 100

Workspace

Metric	Description	Default threshold
Days until token expiry	Days until an access token expires	< 30
Release published	A new release of an Axoniq product	On occurrence
AI credits, Message credits	The credits remaining in the workspace	< 500 AI credits, < 100000 message credits
Message credits exhausted	The workspace has run out of message credits	While exhausted

Metric

Description

Default threshold

Days until token expiry

Days until an access token expires

< 30

Release published

A new release of an Axoniq product

On occurrence

AI credits, Message credits

The credits remaining in the workspace

< 500 AI credits, < 100000 message credits

Message credits exhausted

The workspace has run out of message credits

While exhausted