---
title: "Create alert rules"
description: "Configure Grafana Cloud alerts for Aerospike cluster capacity limits, storage usage, and write stops."
---

# Create alert rules

> For the complete documentation index see: [llms.txt](https://aerospike.com/docs/llms.txt)
> 
> All documentation pages available in markdown.

This page guides you through creating Grafana Cloud alert rules that fire when your cluster approaches capacity limits.

## Key metrics for scaling alerts

| Metric | Warning | Critical | Meaning |
| --- | --- | --- | --- |
| `aerospike_namespace_data_avail_pct` | Below 50% | Below 20% | Storage is filling up |
| `aerospike_namespace_stop_writes` | \- | Equals 1 | Namespace has stopped accepting writes |
| `aerospike_namespace_hwm_breached` | \- | Equals 1 | High water mark breached (memory pressure) |

The metric `memory_free_pct` was removed in Database 7.0. For memory-related alerts, monitor `stop_writes` and `hwm_breached` which indicate when memory thresholds are exceeded.

See [Key metrics to monitor](https://aerospike.com/docs/database/observe/key-metrics) for the complete list of recommended metrics.

## Create a storage usage alert

Create an alert rule from scratch (not from a dashboard panel) for a simpler setup.

1.  In the sidebar, navigate to **Alerts & IRM** > **Alerting** > **Alert rules**.
    
2.  Click **\+ New alert rule**. This creates a fresh alert with no pre-populated queries.
    
3.  Enter a name: `Data usage warning`
    
4.  In the **Define query and alert condition** box:
    
    1.  Select your Prometheus data source (for example, `grafanacloud-yourname-prom`).
        
    2.  Enter the query:
        
        ```plaintext
        100 - aerospike_namespace_data_avail_pct
        ```
        
        This calculates the percentage of storage used.
        
5.  In the **Expressions** box:
    
    1.  The default **Threshold** expression has **Input** set to **A** (your query).
    2.  Set the condition: **IS ABOVE** `50`.
    3.  Click **Preview** to verify. You should see a value (for example, `0` if no data is stored).
    
    This triggers when data usage exceeds 50%, alerting you before storage becomes constrained.
    
6.  In the **Add folder and labels** box:
    
    1.  Click **\+ New folder** and create a folder named `Aerospike`.
7.  In the **Set evaluation behavior** box:
    
    1.  Click **\+ New evaluation group**.
    2.  Enter name: `cluster-health`
    3.  Set **Evaluation interval** to `1m`.
    4.  Click **Create**.
    5.  Set **Pending period** to `1m`.
8.  In the **Configure notifications** box:
    
    1.  Click **View or create contact points**.
    2.  Click **\+ Create contact point**.
    3.  Enter a name: `Aerospike alerts`
    4.  Select integration type: **Email** (or your preferred method).
    5.  Enter your email address.
    6.  Click **Save contact point**.
    7.  Return to the alert rule tab.
    8.  In the **Contact point** dropdown, select **Aerospike alerts**.
9.  Click **Save**.
    

## Create a stop writes alert

1.  Navigate to **Alerts & IRM** > **Alerting** > **Alert rules**.
    
2.  Click **\+ New alert rule**.
    
3.  Enter a name: `Stop writes`
    
4.  In the **Define query and alert condition** box:
    
    1.  Select your Prometheus data source.
        
    2.  Enter the query:
        
        ```plaintext
        aerospike_namespace_stop_writes
        ```
        
5.  In the **Expressions** box:
    
    1.  The default **Threshold** has **Input** set to **A**.
    2.  Set the condition: **IS ABOVE** `0`.
6.  In the **Add folder and labels** box, select the **Aerospike** folder.
    
7.  In the **Set evaluation behavior** box:
    
    1.  Select the **cluster-health** evaluation group.
    2.  Set **Pending period** to **None** (fire immediately).
8.  In the **Configure notifications** box, select **Aerospike alerts**.
    
9.  Click **Save**.
    

## Verify alert rules

1.  In the sidebar, navigate to **Alerts & IRM** > **Alerting** > **Alert rules**.
    
2.  Expand **Grafana-managed** > **Aerospike** > **cluster-health** to see your rules.
    
3.  Verify your rules show state **Normal** (green). This means the condition is not currently met.
    

It may take a moment for the rules to update their state to **Normal** after creation.

::: undefined
-   I’ve created alert rules for storage and stop writes.
:::

[Previous  
Configure Grafana Cloud](https://aerospike.com/docs/cloud/monitor-cloud-with-grafana/step/1/part/2/configure-grafana-cloud) [Next  
Scale with the Cloud API](https://aerospike.com/docs/cloud/monitor-cloud-with-grafana/step/3/part/1/scale-with-api)