Added catch-all rule to system service guan file.

FabricHealer/PackageRoot/Config/LogicRules/SystemServiceRules.guan

@@ -112,4 +112,9 @@ Mitigate(MetricName="FileHandles", OS="Linux", ProcessName=?SysProcName) :- not(
 ## Open File Handles - Linux, Fabric or FabricHost. In these cases, we want a safe (graceful) restart of the Fabric node; not just kill the process, which will restart the node, but not gracefully.
 ## Restart the Fabric node where the offending instance is running.
 
-Mitigate(MetricName="FileHandles", OS="Linux", ProcessName="Fabric") :- TimeScopedRestartFabricNode(2, 08:00:00).
+Mitigate(MetricName="FileHandles", OS="Linux", ProcessName="Fabric") :- TimeScopedRestartFabricNode(2, 08:00:00).
+
+## Generic catch-all.
+## Doesn't matter what the problem is, just restart the service process if it is not Fabric or FabricHost (for hopefully obvious reasons).
+
+Mitigate(ProcessName=?SysProcName) :- not(?SysProcName == "Fabric" || ?SysProcName == "FabricHost"), TimeScopedRestartFabricSystemProcess(2, 08:00:00).

FabricHealer/Repair/RepairExecutor.cs

@@ -600,12 +600,10 @@ namespace FabricHealer.Repair
 
             try
             {
-                // FO provided the offending process id in TelemetryData instance. Chances are good it will still be running.
-                // If the process with this id is no longer running, then we can assume it makes no sense to try to restart it:
-                // Just let the ArgumentException bubble out to the catch.
+                // FO/FHProxy provided the offending process id and (or, in the case of FHProxy) name in TelemetryData instance.
                 if (repairData.ProcessId > 0)
                 {
-                    p = Process.GetProcessById((int)repairData.ProcessId);  
+                    p = Process.GetProcessById((int)repairData.ProcessId);
                 }
                 else // We need to figure out the procId from the FO-supplied proc name.
                 {