> For the complete documentation index, see [llms.txt](https://docs.pharos.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/org-insights/snapshots-and-scans.md).

# Snapshots and Scans

How System Overview builds and refreshes its snapshot, how very large orgs are handled, and what to do when a scan stops or runs long.

System Overview shows a stored snapshot of your org, not a live query. A background scan reads your org and builds that snapshot, either on a schedule or when you click **Scan Now**. Working from a snapshot keeps the page fast and lets Pharos analyze large orgs without the page timing out.

***

## Where the Data Comes From

The page header tells you how fresh the data is.

| Header                  | What It Means                                                                |
| ----------------------- | ---------------------------------------------------------------------------- |
| **Completed at**        | When the snapshot on screen finished building                                |
| **Next Scheduled Scan** | When the next scan starts on its own—or **Not scheduled** if none is planned |

The first time anyone opens System Overview, the first scan starts by itself and the cards fill in when it finishes. On a large org, this can take a while.

All three tabs—Org Overview, Automation, and Security—read the same snapshot. The one exception is the field fill rate under **Unused Fields** on the Automation tab: it is calculated when you pick an object, so it is always current.

***

## Refreshing with Scan Now

To refresh the data once, click **Scan Now** in the header. While the scan runs, the button reads **Scanning…** and is disabled.

* You can leave the page. When you come back or reload, the page picks the running scan back up.
* The tabs keep showing the last complete snapshot until the new one is ready, although with snapshot history off some sections may already show data from the newer scan.

***

## Scan Schedule

By default, a scan runs every 24 hours as part of Pharos's background jobs. **Scan Now** covers one-off refreshes, so change the schedule only if you want a different standing cadence.

To change it, in Salesforce Setup navigate to **Custom Settings** → **Pharos Settings** (managed package custom setting). Edit the org-level record, set **System Overview Snapshot Frequency Hours**, and save.

| Value               | Schedule                               |
| ------------------- | -------------------------------------- |
| **0**               | Off—the header shows **Not scheduled** |
| **1**, **4**, **8** | Every 1, 4, or 8 hours                 |
| **24**              | Every 24 hours (default)               |
| **168**             | Weekly                                 |

{% hint style="warning" %}
**Background Load:** Each scan reads across your whole org, so a shorter cadence adds background load. On very large orgs, keep the default of 24 hours or use 168 (weekly).
{% endhint %}

**Not scheduled** can also mean that Pharos's background scheduler isn't running. Check the **Scheduled Jobs** list on the [Pharos Running User](https://docs.pharos.ai/configuration-guide/advanced/pharos-running-user.md) page.

***

## Snapshot History

With snapshot history on, Pharos keeps each scan, and the **Trend** buttons on the cards chart how values change from scan to scan (a single point until a second scan has completed). With it off, each scan replaces the last, and Trend shows a single point. Snapshot history is off unless Pharos turns it on for your org. If you'd like it, contact Pharos support.

***

## Very Large Orgs

System Overview is designed for objects with very large record volumes. Rather than reading every record, the scan uses approximate or bounded record counts, and Unused Fields calculates fill rates from a sample.

### Record Counts May Be Approximate

Where Salesforce's record-count statistics are available, the counts on the **Record Count by Object** card (Automation tab) come from them. These are approximate counts and can differ from an exact count, so the card carries the caption "Approximate — from Salesforce record-count statistics".

Where those statistics aren't available, Pharos counts the records itself: exactly up to 50,000, and as **50,000+** above that, meaning the object has at least 50,000 records. If no count on the card came from those statistics, the card shows no caption. If no count can be read for an object, it shows **—**.

### Field Fill on Large Objects

Field fill rates are exact for objects with up to 50,000 records. Above that, they are based on a sample of the newest records. See [Automation → Unused Fields](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/org-insights/automation.md#unused-fields) for how to read them.

### A Section Too Large to Open

On some orgs, part of the stored snapshot is too large for the page to open. When that happens, a notice at the top of the affected tab names the part that could not be opened, and the rest of the page keeps working. If the notice is still there after later scans, contact Pharos support.

***

## Long Scans

While a scan runs, the page keeps checking on it, less often as time passes.

* **After about 10 minutes**, the page shows a note naming the part of the scan in progress and saying that large orgs can take a while. This isn't an error.
* **After about 90 minutes**, the page stops checking and asks you to reload later. The scan carries on in the background, and reloading picks it back up.

{% hint style="info" %}
**Stopped Scans:** A stopped scan shows as soon as Pharos detects it. Until then, a step that has stopped can still look as if it is running—for up to about an hour—because Salesforce still reports it as running.
{% endhint %}

***

## When a Scan Stops

A scan can stop part-way, for example when Salesforce ends one of its background jobs. What you see depends on whether an earlier complete snapshot exists.

### No Earlier Complete Snapshot

A panel headed "The last System Overview scan did not finish." replaces the tabs. The header, with **Scan Now**, stays in place. The panel shows:

* **Where and when it stopped**—the part of the scan, such as "Objects (automation and record counts)", and the time. If the part isn't known, the panel shows when the scan started instead.
* **The reason** Salesforce recorded. If Salesforce no longer has the job record, the panel says so: Salesforce keeps job records for 7 days.
* A **Scan again** button that starts a new scan.

### An Earlier Complete Snapshot Exists

The page keeps showing the earlier snapshot. A banner above the tabs names the part of the scan where the latest scan stopped, when, and the reason, along with the date of the snapshot you're looking at. With snapshot history off, the banner adds that some sections may already come from the newer scan.

### What Happens Next

Pharos does not restart a stopped scan just because you open the page. Click **Scan again** or **Scan Now**, or wait for the next scheduled scan, which runs at its usual time.

When Pharos can identify the stopped job, it also records the stopped scan as a log in the **Pharos Error** category, with the part of the scan, the reason, and the times.

| Situation                         | What to Do                                                                                                                                                                                             |
| --------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
| The scan stopped once             | Click **Scan again**                                                                                                                                                                                   |
| It stops again in the same part   | Contact Pharos support with the part name, the reason, the time, and the Pharos Error log if there is one                                                                                              |
| The header says **Not scheduled** | Check the frequency setting under [Scan Schedule](#scan-schedule) and the Scheduled Jobs on the [Pharos Running User](https://docs.pharos.ai/configuration-guide/advanced/pharos-running-user.md) page |

{% hint style="success" %}
**Next Steps:** With current data in place, start with the [Org Overview](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/org-insights/org-overview.md) for a baseline, then explore [Automation](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/org-insights/automation.md) and [Security](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/org-insights/security.md) for deeper analysis.
{% endhint %}
