[SalesForce] Platform Event Apex trigger limits exception – where does it go

If one publishes a Platform Event, the consuming trigger executes as Automated Process user. If that consuming trigger exceeds CPU limits, no notification is sent to the developer(s).

Even though Apex Exception Email is configured to route to me

Even though my user is configured to get Apex Warning Emails

How does one discover that there was a limits exception so action can be taken?

Platform Event Foo__e

Apex Trigger

trigger FooTrigger on Foo__e (after insert) {
    Util.sleep(30); // method that does a CPU sleep of ~arg secs 
}

DeveloperConsole

EventBus.publish(new Foo__e());

I have no idea where to find the limits exception email or other indication of failure; querying EventBusSubscriber yields no information re: errors of this sort.

Support ticket filed

UPDATE:

Lesson to us all: Since a Limits failure fails silently when running
user = Automated Process, and since your platform event trigger is
bulkified; if one event in your batch is, say, exceptionally CPU or
heap intensive, the other batch members will fail (silently) as
"collateral damage" and be undetected until you do a database audit.
Adrian Larson's answer is a practical design way out of this.

Ask me how I know 🙁

Best Answer

My plan to work around this limitation would be roughly as below.

Configuration

Use a logging object to track each Platform Event you process in your trigger.
- Make sure this logging object has no required fields nor validation rules.
Add a Text (18) field to track Job_Id__c.
- You can later use this field to query AsyncApexJob.
Add a Text field to track the Job_Status__c.
Add a TextArea field to track the Job_Error_Message__c.

Code Changes

Move your core logic to a Queueable.
From your subscriber trigger, fire this async job so any steps which can fail will take place in a separate transaction.
From your subscriber trigger, insert a record into your log object.
On your log record, set the Job_Id__c field.
Set up a scheduled batch which iterates over any log records whose Job_Status__c is in (null, 'Holding', 'Preparing', 'Processing').
- Match each Job_Id__c up to the corresponding AsyncApexJob.
- Map Status to Job_Status__c.
- Map ExtendedStatus to Job_Error_Message__c.

Related Solutions

[SalesForce] Do Platform Event Triggers process synchronously? Do they get a fresh set of limits

Are the subscription triggers processed synchronously or asynchronously?

Aynchronously

Do the subscription triggers get separate limits from the publishing context?

Yes, separate limits from the publishing context

If the limits are separate, does each subscription trigger get its own set?

Each Apex subscription trigger executes in its own transaction with independent limits.

To back these claims up I create a test platform event with two Apex trigger subscriptions. I then executed a publish for 300 events via anonymous Apex.

There were 5 independent transaction logs generated. One for the Tooling API execute anonymous, followed by 4 "Platform Event Trigger" operations. Historically each log entry has always represented a separate transaction (with their own limit). I'm assuming the same applies here. I'll add some more limit logging to back that up.

Sequence of logs for my second test attempt:

Operation /services/data/v40.0/tooling/executeAnonymous/ to publish the 300 events. Timestamp: 7:57:09 PM
Operation Platform Event Trigger 2nd Apex trigger with CPU Limit at 36 with last DEBUG message. Trigger.new size of 300. Timestamp: 7:57:11 PM
Operation Platform Event Trigger 1st Apex trigger with CPU Limit at 23 with last DEBUG message. Trigger.new size of 200. Timestamp: 7:57:11 PM
Operation Platform Event Trigger 1st Apex trigger with CPU Limit at 12 with last DEBUG message. Trigger.new size of 100. Timestamp: 7:57:12 PM

Note how my first test split 200 events to the first event trigger transaction and then 100 events to the latter transactions. Yet in my second test with the same triggers and same number of published events the second trigger only fired once for the full 300 records. Important note on potential Trigger.new size for platform events.

It can potentially go all the way up to 2000 events in one trigger transaction. Source

[SalesForce] Queueable Apex not working in Platform Event Trigger

After struggling with this for longer than I'd like to admit, I found a method that enables us to test Platform Event Triggers that enqueue Queueable jobs. The magical method is:

Test.getEventBus().deliver();

What you need to do is call Test.getEventBus().deliver() after you have called EventBus.publish(platformEvents). Just make sure you do this between Test.startTest() and Test.stopTest()

The gotcha is that YOU MUST call Test.getEventBus().deliver() before Test.stopTest()

Here's some code to illustrate:

Test.startTest();

// Publish a platform event that will enqueue a `Queueable` job
EventBus.publish(new Some_Platform_Event__e(Foo__c = 'bar'));

// Deliver the platform event RIGHT NOW
Test.getEventBus().deliver();

// Let's see what happened
List<AsyncApexJob> asyncApexJobs = [
    SELECT Id, Status
    FROM AsyncApexJob
    WHERE JobType = 'Queueable'
];

// We have 1 Queueable job that has been Queued
System.assertEquals(1, asyncApexJobs.size());
System.assertEquals('Queued', asyncApexJobs[0].Status);

Test.stopTest();

// Now let's see what happened after stopping the test
asyncApexJobs = [
    SELECT Id, Status
    FROM AsyncApexJob
    WHERE JobType = 'Queueable'
    AND Id = :asyncApexJobs[0].Id
];

// The Queueable has completed!
System.assertEquals('Completed', asyncApexJobs[0].Status);

Best Answer

Related Solutions

[SalesForce] Do Platform Event Triggers process synchronously? Do they get a fresh set of limits

[SalesForce] Queueable Apex not working in Platform Event Trigger

Related Topic