I have noticed that sometimes Additional Polling Engines could have an issue with saving new SNMP responses to the database.
There is a dedicated Database Sync value, but it is much better to check, if the samples are really in the raw data of the database.
By default, alert is checking samples from last 30 minutes and it will try to restart the Collector Service (assuming that you are monitoring "SolarWinds Collector Service").
If you do not have SAM module, please remove this action from Triggers.
Also, in order for the alert to work properly, you need to add APEs as nodes, at least via ICMP or External.
SELECT Nodes.Uri, Nodes.DisplayName FROM Orion.Nodes AS Nodes
JOIN Orion.OrionServers s ON Nodes.Caption = s.HostName
JOIN Orion.Engines e ON s.HostName = e.ServerName
JOIN (
SELECT
N.Engine.DisplayName AS [DisplayName]
,N.Engine.Uri AS [Uri]
,count(N.NodeID) AS [All]
,a.SNMPproblems
,a.SNMPproblems/count(N.NodeID) AS [Perc]
FROM Orion.Nodes N
JOIN
(
SELECT
N.Engine.DisplayName AS [Engine]
,count(N.NodeID) AS [SNMPproblems]
FROM Orion.Nodes N
-- Checking if there is a new sample in the last 30 minutes
WHERE N.LastSystemUpTimePollUtc < AddDate('minute', -30, GETUTCDATE())
AND N.Status = 1
GROUP BY N.Engine.DisplayName
) a ON a.Engine = N.Engine.DisplayName
WHERE N.Status = 1
GROUP BY N.Engine.DisplayName, a.SNMPproblems, N.Engine.Uri
ORDER BY N.Engine.DisplayName
) b ON b.DisplayName = e.ServerName
WHERE (s.ServerType = 'AdditionalPoller') and b.Perc = 1