> For the complete documentation index, see [llms.txt](https://cleyrop.gitbook.io/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://cleyrop.gitbook.io/docs/documentation-fr-en/administration/surveiller-les-ressources.md).

# Monitor resources

The platform resources (excluding core resources) are divided into two nodes:

* **Compute** : Jupyter sessions and execution clusters for Dataflows.
* **Apps** : Applications and DevSpaces.

Resource monitoring allows **Platform Managers** to view the consumption of the different services. These indicators help anticipate saturation, identify consumption spikes, and intervene before a limit is reached.

***

## Compute resources (cluster & Codelab)

**To access monitoring**

1. Go to **Administration > Monitoring**
2. Click on the **Resources**
3. You can select the desired period using the selector at the top right

<figure><img src="/files/20a9a64eec4aba0e43e05077478de791ada73f52" alt=""><figcaption></figcaption></figure>

**Understanding the charts**

The page displays two charts: **CPU** and **RAM (in GB)**, showing how consumption evolves over time for PySpark/Python clusters and Jupyter sessions.

<table><thead><tr><th width="236.7421875">Metric</th><th width="157.33984375">Color</th><th>Meaning</th></tr></thead><tbody><tr><td>Average consumption</td><td>Green</td><td>Average consumption of Dataflows and Codelab sessions over the displayed period</td></tr><tr><td>Maximum consumption</td><td>Beige</td><td>Consumption peak reached over the displayed period</td></tr><tr><td>Limit</td><td>Dark blue</td><td>Maximum resources currently available for clusters and Jupyter sessions</td></tr></tbody></table>

The Limit reflects the **resources available for clusters and Jupyter sessions** on the platform. It may change over time depending on resource availability. It cannot be modified from the interface — contact Cleyrop support if you need to know more about your available capacity.

{% hint style="danger" %}
**Warning** : this limit is a capacity **aggregated total across the entire compute node pool**. It does not guarantee that a resource can be deployed: if no machine in the pool individually has sufficient resources (CPU/RAM) to host a cluster or a session, deployment will fail, even if the total available seems sufficient. An improvement to the monitoring is planned to provide better visibility into this distribution by machine.
{% endhint %}

**Example** **reading**

<figure><img src="/files/91f1588b16eae5f6e69c44be107bb0ec55371ee5" alt=""><figcaption></figcaption></figure>

In this example:

* The CPU limit is 8 CPU for most of the period — it was 16 CPU at the start of the period, reflecting a different available capacity for clusters and Jupyter sessions at that time.
* On May 5, maximum consumption reaches 8.04 CPU, nearly hitting the limit, while the average remains at 6.74 CPU that day.
* Between May 3 and May 6, a significant spike is also visible in RAM.

## Resources used by The Factory &#x20;

**To access monitoring**

1. Go to **The Factory  > Monitoring**
2. Click the button at the top right **Used resources**

<figure><img src="/files/b7bcb67050131b554c4cdffa24acc19dd4cbfe2d" alt=""><figcaption></figcaption></figure>

The panel shows, for all elements successfully deployed and launched in your environment, the sum:

<table><thead><tr><th width="261.89453125">Metric</th><th>Meaning</th></tr></thead><tbody><tr><td>CPU Min. / Max.</td><td>CPU resource range configured when instances are deployed</td></tr><tr><td>RAM Min. / Max.</td><td>RAM resource range configured when instances are deployed</td></tr></tbody></table>

These values correspond to the resources **reserved for the configuration**, and not to consumption observed in real time.

{% hint style="info" %}
**The totals include native Cleyrop Tools but do not include .** The resources displayed in the panel aggregate all Apps, Tools, and images deployed on the environment, including the platform's native components (dataset search tools, corpus search...).
{% endhint %}

***

**Best practices**

* If the **maximum consumption** regularly approaches the **limit**, contact your Customer Success Manager to assess a capacity increase.
* Remember to stop unused instances in The Factory  to free up reserved resources.