# Operations Center

> What the Operations Center shows. The verdict, the Needs attention list, the fleet, the data pipeline and the events of the last changes.

Source: https://proxus.io/docs/monitor/operations-center

The **Operations Center** answers three questions in order: is anything wrong, where, and what should you do. It
is the first item of the **Monitor** menu. The page refreshes by itself; **Live** and **Updated** in the top right
corner show when it last did.

The page keeps normal things grey. Only something that needs you gets a colour and a place in the list.

*Screenshot: The Operations Center with the verdict at the top, one open alarm and one note in Needs attention, and the fleet with three sites*

1. **Verdict**: one line that sums up the hub. **All systems within limits** when nothing needs you, otherwise
   the number of issues and what they are.
2. **Needs attention**: the open alarms and system problems, worst first.
3. **Counts**: sites, gateways, devices and delivery targets, each with a short note when something is wrong.
4. **Fleet**: one row per site with one cell per device.
5. **Data flow map**: opens a full-screen map from each gateway to its devices and their targets.

## Verdict

When nothing needs attention, the verdict lists how many devices are sending data, the values per second and the
latency, that there are no open alarms, and how many notes are below. When something needs attention, it says how
many issues there are and lists them. Its colour follows the worst item: amber for a warning, red for a critical
issue.

## Needs attention

The list holds real alarms and problems the hub finds by itself, in one list. Critical items come first; within the
same level, the oldest comes first, because it has waited longest. Each item has a level badge (**CRIT**, **WARN**
or **NOTE**), a title, where it is, how long ago it started and the buttons that fit it.

| Item | When it appears | Buttons |
| --- | --- | --- |
| An open alarm | A rule or a pipeline raised an alarm that nobody has acknowledged yet. The age shows **unacknowledged**. | **Acknowledge** takes the alarm. **Open** opens it on the alarm page. |
| **Gateway … is silent** | An edge gateway stopped reporting. Its devices are not counted again as devices without data. | **Probe** sends a test message to the silent gateways and shows which ones answer. **Gateways** opens the gateway list. |
| **… devices are not sending data** | Enabled devices that sent no data, for a reason other than a silent gateway. It is a **NOTE** while fewer than 5% of the devices are affected, and a warning above that. | **Check** lists the devices without an active stream and the last error of each. **Devices** opens the device list. |
| **No device is sending data** | There are devices, but none of them sends data. | **Check**, **Devices** |
| Critical log entries, or errors logged in the last minute | At least one critical log entry, or five or more errors, in the last minute. | **Open logs** opens **System Logs**. |
| **License capacity near its limit** or **exceeded** | A gateway's outbound data points per second are close to, or above, what the license allows. Above the license, outbound data points are dropped. | **Gateways** |
| **Streaming devices have no delivery target** | Devices send data, but no device has a target, not even the historian (**SYSTEM INTERNAL DB**). Values are visible live, but are not forwarded or stored. | **Targets** opens the outbound targets. |

**Check** and **Probe** only look; they do not restart or change anything. Their result stays under the item until
the page is opened again.

Below the list, **… acknowledged, in follow-up** opens the alarms someone has acknowledged but not closed yet, with
who took each one and when. Select one to open it. **All alarms** opens the **Alerts** list.

## Fleet

| Count | What it shows |
| --- | --- |
| **Sites** | Sites with no issue, out of all sites. **No site assigned** is listed below but not counted as a site. |
| **Gateways** | Connected gateways, out of all gateways. |
| **Devices** | Devices sending data, out of all devices. |
| **Delivery** | Active outbound targets. |

Select a count to open the matching list.

Below the counts, each site has a row: its gateways and devices, a cell per device and a short state such as
**Normal** or **1 device without data**. A grey cell is a device sending data, an amber cell a device without
data, and a striped cell a device behind a silent gateway. A site with very many devices shows a bar with the share
of devices sending data instead of cells. With more than eight sites, the panel shows one tile per site and lists
only the sites with problems.

**Data flow map** opens a full-screen tree: each gateway, the devices it reads and the targets their data goes to.
Select **Close** to go back.

## Data pipeline

*Screenshot: The Data pipeline section with ingest, processing and delivery, the values per second chart with an alarm marker, and the Events list below*

1. **Ingest**, **Processing** and **Delivery**: the three stages of the data. Ingest shows the values per second,
   the latency, how many devices send data and the protocols in use. Processing shows the active rules and
   functions and the open alarms. Delivery lists up to five active outbound targets and how many devices each
   one receives.
2. The chart: values per second and latency over the time shown next to the section title (up to the last hour).
   A dashed line marks the moment an alarm was raised. The chart appears after the first minute of data.
3. **Events**: see below.

## Events

**Events** lists the latest changes of state: alarms raised, acknowledged, resolved or ignored, gateways that went
silent or reconnected, devices that stopped or resumed sending data, license capacity changes and error bursts.
Each line has the time, the kind (**Alarm**, **Connectivity**, **Delivery**, **Capacity** or **System**), the
message and, for alarms, who changed it. **Logs** opens **System Logs**.

> **Note:**
Changes that were not stored as alarms, such as a gateway going silent, are only listed from the moment the page
was opened. If nothing changed since then, the list says so.

## What's next

- [Find what stopped](https://proxus.io/docs/monitor/find-what-stopped): follow a warning to the device and its error.
- [Acknowledge and resolve an alert](https://proxus.io/docs/monitor/acknowledge-an-alert)
- [Topology](https://proxus.io/docs/monitor/topology): the same health, drawn as a map.