Automatic monitoring

The checks work with fixed rules and statistics over the readings. They are set per gateway in Gateway settings, in the Automatic monitoring block. An owner, admin or technician can change them. With read-only access you cannot change anything.

What the checks report

The table shows what each check reports and which kind of register it applies to.

The limit of Modbus device not reporting is 5 times the polling interval of the Modbus device, and at least 15 minutes. If a Modbus device has never delivered a value, this check reports nothing. If the whole gateway drops out, the Gateway offline alert rule reports that. To find the cause, use Modbus device shows no values.

States of a check

Every check is set to Off, Trial run or On. On Trial run it records what it would report, without an alert or email. The Last 7 days column counts how often that happened. On On it creates alerts and sends email to the recipients of the gateway.

Alerts from monitoring appear under Alerts with Automatic after the title. In the All sources list, choose Automatic monitoring to see only these alerts. Otherwise they work as described in How alerts work.

If you switch a check from On to Trial run or Off, the portal resolves its open alerts. It asks for confirmation first.

Default alerts on a new gateway

A gateway that you link with Add a gateway gets these alert rules straight away:

• Gateway offline, which reports when the gateway goes offline.
• The five checks, set to On.
• One rule per register of the kind Fault code. It gives a critical alert if the register reports anything other than 0 for two readings in a row.

The email from these rules goes to the recipients you set with Choose who receives alert emails. A Modbus device that you add later falls under the checks automatically. It gets fault code rules only if the gateway already has the default alerts. A gateway without default alerts never gets them automatically.

Apply default monitoring to an existing gateway

Loading the walkthrough.

On a gateway without default alerts, you add them with Apply default monitoring. If the gateway already has them, the date on which they were applied appears under the Automatic monitoring heading.

1. Open the gateway list

   Open Gateways in the menu.

2. Open the gateway

   Click the gateway.

3. Open the settings

   Click the pencil at the top right (Gateway settings).

4. Start applying

   Under Automatic monitoring, click Apply default monitoring. The Apply default monitoring? window shows which alert rules the portal creates if they do not exist yet.

5. Choose email

   If you want to see these alerts only in the portal, clear Send email to the recipients of this gateway.

6. Choose a trial run

   If the checks must be on straight away, clear Let automatic monitoring do a trial run first (recommended for a gateway that is already running).

7. Add the rules

   Click Apply. The portal tells you how many alert rules were added.

The portal creates only the rules that are missing. Existing rules, such as your own offline rule, stay as they are with their state. For email that you clear here, an owner or admin turns it on per rule later in Settings > Notifications, under Outgoing mail.

Turn on a check after the trial run

After the trial run, you set each check to On separately.

1. Open the settings

   On the gateway page, click the pencil at the top right (Gateway settings).

2. Expand the block

   Under Automatic monitoring, click Show. The checks appear with their state in the Status column and the count in Last 7 days.

3. Choose the state

   In the Status column, choose On for the check.

4. Confirm turning on

   Click Turn on. From now on the check creates alerts.

If a check has no alert rule yet, it is set to Off and Severity and email is missing from its row. If you choose Trial run or On, the portal creates the rule with the default settings. If a check fires too often, see Alert does not fire, fires too often or stays open.

Kind and learned normal per register

The Kind column under Registers decides which checks a register takes part in. The portal derives the kind itself and then shows, for example, Automatic: Measurement. If that is wrong, choose a kind yourself, such as Counter or Fault code. Monitoring uses the new kind from the next nightly calculation.

Registers with the kind State, Setting, Bit field or Unknown, outage only fall only under Modbus device not reporting. If you later set a register to Fault code, click Apply default monitoring again. The register then gets its fault code rule.

Deviation from learned normal first learns what is normal for each hour of the day per register. This takes at least 14 days, and until then this check reports nothing for that register. The Learned normal column shows when a register is expected to be ready. Registers for voltage, current, frequency, power factor and apparent or reactive power get no learned normal, and neither does a value that changes too little per hour.

Exclude a Modbus device or register

Excluding removes one Modbus device or register from a check. The portal saves every choice straight away. You exclude a Modbus device by turning off the Monitor switch under Outage per Modbus device. You exclude a register per check.

1. Expand the block

   Under Automatic monitoring, click Show.

2. Open the choice

   Under Registers, in the Exclude column, click the choice for the register. Without an exclusion it says None.

3. Select the check

   Under Leave out of, select the check that the register must skip.

If you cannot select a check, First set this monitor to trial run or on. appears below it. Choose that state first.

Set severity and email per check

Every check is an alert rule with its own severity, cooldown and recipients. The table shows the settings each check starts with.

If a kWh counter for consumption from or feed-in to the grid runs backwards, the Counter fault alert is always critical.

1. Expand the block

   Under Automatic monitoring, click Show.

2. Open the rule

   In the row of the check, click Severity and email. The Edit alert rule page opens.

3. Choose the severity

   Under Severity, click Info, Warning or Critical.

4. Enter the cooldown

   In Cooldown, enter the number of minutes.

5. Save the changes

   Click Save changes.

You enter extra addresses in Email recipients. That field and the choice under Send work as in Create an alert rule.