Some times agents start to stop collecting data without stopping completely . this is often when client are overloaded or management servers are for a while unavailable. In my case most times disk space counters stopped working as on of the first roles. So check this counter frequently, like every week or after maintenance jobs.
It can be easy done by using a query towards ops db.
First create a view
CREATE VIEW [dbo].[50beansDiskFreeSpace] AS SELECT TOP (10) PERCENT bme.Path, ps.PerfmonInstanceName, pdav.SampleValue, pdav.TimeSampled, dbo.MaintenanceMode.IsInMaintenanceMode FROM dbo.PerformanceDataAllView AS pdav WITH (NOLOCK) INNER JOIN dbo.PerformanceSource AS ps WITH (NOLOCK) ON pdav.PerformanceSourceInternalId = ps.PerformanceSourceInternalId INNER JOIN dbo.Rules AS r WITH (NOLOCK) ON ps.RuleId = r.RuleId INNER JOIN dbo.BaseManagedEntity AS bme WITH (NOLOCK) ON ps.BaseManagedEntityId = bme.BaseManagedEntityId INNER JOIN dbo.MaintenanceMode ON bme.BaseManagedEntityId = dbo.MaintenanceMode.BaseManagedEntityId WHERE (r.RuleName LIKE N'%LogicalDisk.FreeMB%' OR r.RuleName = N'Microsoft.Windows.Server.ClusterDisksMonitoring.ClusterDisk.Monitoring.CollectPerfDataSource.FreeSpaceMB') AND (pdav.TimeSampled = (SELECT MAX(TimeSampled) AS Expr1 FROM dbo.PerformanceDataAllView WHERE (PerformanceSourceInternalId = pdav.PerformanceSourceInternalId))) ORDER BY pdav.TimeSampled GO
Now run the view . On top you will find all disks in Maintenance mode. It is clear they did not collect any data within the last few minutes . After that follow all servers with a partly disabled agent. on the bottom you will find clients having an TimeSampled within an few minutes. They are working well.