# Health

> How your Dedicated Cloud is doing — disk, memory and connection usage, security and performance checks with Fix with AI, and logs.

Source: https://support.estage.com/cloud-health
Part of the ESTAGE Support help center (https://support.estage.com). Full knowledge base: https://support.estage.com/llms-full.txt

---

The **Health** tab shows how your backend is doing, what to fix, and what it logged. Open it
when your site is slow, a save fails, or you want to check the database is set up safely.

Find it under **Health** in the [Cloud workspace](https://support.estage.com/cloud#the-cloud-workspace). The refresh
button at the top reloads everything on the tab.

## What needs attention

Warnings appear at the top of the tab when something is close to a limit:

| Warning | What it means for your site |
| --- | --- |
| **Your database is read-only** | The disk is full, so sign-ups, orders and every other save fail. The disk is enlarged automatically, which can take a few hours. Deleting data you no longer need frees space sooner. |
| **The database disk is …% full** | Shown from 90%. The disk grows as it fills, but a sudden burst can outpace it, and a full disk makes the database read-only. |
| **Memory is almost full** | Queries slow down. |
| **Close to the connection limit** | When no connections are left, new requests from your site wait or fail. |
| **The processor is busy** | If it stays busy, pages that read the database get slower. |

For memory, connections and processor, the project owner sees **Bigger plan**, which opens the
plan picker — see [Changing your plan](https://support.estage.com/cloud#changing-your-plan). It isn't shown on the top
plan.

A full or nearly full disk also shows as a badge in the workspace header, such as
**Read-only: disk full**, and the project owner gets an email about it.

## Usage

| Tile | Shows |
| --- | --- |
| **Disk** | How full the database disk is, used of total, and **read-only** when it is |
| **Database** | The size of your tables, indexes and users |
| **Memory** | How much of the backend's memory is in use |
| **Processor** | The current load and the number of cores |
| **Connections** | Connections open to the database, out of the maximum |

The bars turn amber at 80% and red at 90%.

## Security and performance checks

Two cards check your database for common problems:

- **Security** — for example, tables your site can read or write without the right
  [access rules](https://support.estage.com/cloud-database). When everything passes, it reads *"No security problems
  found."*
- **Performance** — for example, a table without a primary key, or an index that is missing
  or never used. When everything passes, it reads *"Nothing to speed up."*

Each finding has a level — **Fix**, **Check** or **Tip** — and the number of places it applies.
Select a finding to see what it means and where it applies. Select the refresh button on a card
(**Check again**) to run its check again.

### Fix with AI

**Fix with AI** on a finding opens the builder chat and sends Genesis a request to fix it. Genesis
reads your live database first, makes the fix as a database change, keeps your site's existing
reads and writes working, and tells you in plain words what it changed.

The fix is a regular chat turn and uses credits like any other turn. The change is listed in
[Backups → Change history](https://support.estage.com/cloud-backups#change-history).

## Logs

The **Logs** card shows recent lines from one part of the backend:

1. Pick the source: **API requests**, **Database**, **Sign-in**, **Functions**, **Storage** or
   **Realtime**.
2. Pick the period: **Last 15 minutes**, **Last hour** or **Last 24 hours**.
3. To filter, type in **Search the logs** and press Enter.

For one function's logs, open it in the [Functions](https://support.estage.com/cloud-functions) tab.

## Things worth knowing

- **Anyone who can open the Cloud workspace can see Health.** Only the plan controls are limited
  to the project owner.
- **A read-only database fixes itself.** You don't have to do anything, but deleting data you no
  longer need brings saves back sooner.
- If a check or the logs can't be read, the card says *"Try again in a minute."*
