2 Minute Plugin Time Out Unhandled Exception - dynamics-crm

Question

I'm trying to simulate a 2 minute time out on my Plugin by using Thread.Sleep(120000) to generate a Time Out but when that exception arises my try catch can't seem to catch them, even the finally block is skipped.
How would i properly catch it so that i can perform on creating a case record on this error?
I've tried using different catch exceptions to no avail. I've even tried profiling the plugin but it also won't profile the plugin due to the error.
protected override void ExecuteCrmPlugin(LocalPluginContext localContext) {
IPluginExecutionContext context = localContext.PluginExecutionContext;
IOrganizationService service = localContext.OrganizationService;
ITracingService tracer = localContext.TracingService;
try {
if (context.InputParameters.Contains("Target") && context.InputParameters["Target"] is Entity) {
Entity entity = (Entity)context.InputParameters["Target"];
if (context.MessageName.ToLower() != "create")
return;
if (entity.LogicalName.ToLower() == Contact.EntityLogicalName) {
tracer.Trace("Sleep for 2 min");
Thread.Sleep(120000);
}
}
}
catch (System.ServiceModel.FaultException<System.TimeoutException ex) {
throw new InvalidPluginExecutionException("An error occurred in the plug-in.", ex);
}
catch (System.ServiceModel.FaultException<Microsoft.Xrm.Sdk.OrganizationServiceFault ex) {
throw new InvalidPluginExecutionException("An error occurred in the plug-in.", ex);
}
catch (TimeoutException e) {
throw new InvalidPluginExecutionException("A timeout has occurred during the execution of the plugin.", e);
}
catch (FaultException ex) {
throw new InvalidPluginExecutionException("Err occurred.", ex);
}
catch (Exception ex) {
tracer.Trace(ex.ToString());
throw;
}
finally {
tracer.Trace("Finally");
}
}
In the Base Class I also have the same catch blocks.
The error is:
Unhandled exception:
Exception type: System.ServiceModel.FaultException`1[Microsoft.Xrm.Sdk.OrganizationServiceFault]
Message: An unexpected error occurred from ISV code. (ErrorType = ClientError) Unexpected exception from plug-in (Execute): MyPlugin.Plugins.PreCreateContact: System.TimeoutException: Couldn’t complete execution of the MyPlugin.Plugins.PreCreateContact plug-in within the 2-minute limit.Detail:
<OrganizationServiceFault xmlns:i="http://www.w3.org/2001/XMLSchema-instance" xmlns="http://schemas.microsoft.com/xrm/2011/Contracts">
<ActivityId>397e0f4c-2e16-43ea-9368-ea76607820a5</ActivityId>
<ErrorCode>-2147220956</ErrorCode>
<ErrorDetails xmlns:d2p1="http://schemas.datacontract.org/2004/07/System.Collections.Generic" />
<HelpLink i:nil="true" />
<Message>An unexpected error occurred from ISV code. (ErrorType = ClientError) Unexpected exception from plug-in (Execute): MyPlugin.Plugins.PreCreateContact: System.TimeoutException: Couldn’t complete execution of the MyPlugin.Plugins.PreCreateContact plug-in within the 2-minute limit.</Message>
<Timestamp>2019-07-17T00:49:48.360749Z</Timestamp>
<ExceptionRetriable>false</ExceptionRetriable>
<ExceptionSource>PluginExecution</ExceptionSource>
<InnerFault i:nil="true" />
<OriginalException>System.TimeoutException: Couldn’t complete execution of the MyPlugin.Plugins.PreCreateContact plug-in within the 2-minute limit.
at System.Runtime.ExceptionServices.ExceptionDispatchInfo.Throw()
at Microsoft.Crm.Sandbox.SandboxAppDomainHelper.Execute(IOrganizationServiceFactory organizationServiceFactory, Dictionary`2 sandboxServices, String pluginTypeName, String pluginConfiguration, String pluginSecureConfig, IPluginExecutionContext requestContext, Boolean enablePluginStackTrace, Boolean chaosFailAppDomain, String crashOnPluginExceptionMessage)
at Microsoft.Crm.Sandbox.SandboxAppDomainHelper.Execute(IOrganizationServiceFactory organizationServiceFactory, Dictionary`2 sandboxServices, String pluginTypeName, String pluginConfiguration, String pluginSecureConfig, IPluginExecutionContext requestContext, Boolean enablePluginStackTrace, Boolean chaosFailAppDomain, String crashOnPluginExceptionMessage)
at Microsoft.Crm.Sandbox.SandboxWorker.<>c__DisplayClass3_0.<Execute>b__0()</OriginalException>
<TraceText>
Entered MyPlugin.Plugins.PreCreateContact.Execute(), Correlation Id: 0c2b0dd3-d27c-46ea-a7e2-90c0729b326e, Initiating User: 61e01dfa-668a-e811-8107-123456
</TraceText>
</OrganizationServiceFault>

Answer 1

I guess once the plugin hits the 2-minute timeout, our ability to process anything is over. And that makes sense - if the system allowed us to run a catch block after the 2-minute timeout that catch block could run for another minute, or five.
One approach I've taken to allow a plugin or custom workflow process something that might take longer than the 2-minute timeout is to use a cancellation token to gracefully shut down before the timeout. Rather than catch the platform's timeout exception, you can throw your own at 115 seconds and cancel gracefully.
I've even gone so far as to preserve the state of the execution as XML stored in an annotation and retrigger the plugin up to the infinite loop limit of 7 times. And beyond that I've created four processes that each recur every hour, staggered by 15 minutes - i.e. 15, 30, 45, and 60 minutes after the main process. (A process can recur hourly without triggering an infinite loop exception). These processes catch any jobs that have preserved their state after 7 runs and retrigger them for another 7. Using this method I have had workflows complete calculations that took hours.
Mind you, this could be considered abuse of the asynchronous processing system. I'll admit it's a stretch, but the use case needed to handle big calculations entirely inside of Dynamics. Of course the standard approach to "escape" the sandbox timeout is to move processing to an external service (e.g. Azure API, Azure Functions, Flow, Logic Apps, etc.)