Examkin

Question

Difficulty: HardImplement Azure Event Hubs Solutions

An organization runs a C# (.NET) microservice to ingest IoT telemetry. The service utilizes the `Azure.Messaging.EventHubs.Processor` library and deploys multiple instances of `EventProcessorClient` that share a common Azure Blob Storage container as a checkpoint store. During peak loads, you observe partition thrashing where consumer instances continuously claim and release partition ownership from each other, resulting in excessive duplicate processing. Upon reviewing the client configuration, you find the following setup:

csharp
var options = new EventProcessorClientOptions
{
LoadBalancingUpdateInterval = TimeSpan.FromSeconds(15),
PartitionOwnershipExpirationInterval = TimeSpan.FromSeconds(10)
};

Which of the following actions should you perform to resolve the partition thrashing?

  1. A
    Implement a custom BlobCheckpointStore that manually acquires and releases exclusive leases on the individual partition blobs using a BlobLeaseClient to prevent concurrent ownership modifications.
  2. Increase the PartitionOwnershipExpirationInterval to a value that is significantly greater than the LoadBalancingUpdateInterval, such as 45 seconds.Answer
  3. C
    Change the authentication of the EventProcessorClient from a system-assigned managed identity to a user-assigned managed identity to prevent instances from sharing lease credentials.
  4. D
    Migrate the event consumption workload to an Azure Service Bus Queue and configure the client to receive messages in ReceiveAndDelete mode.

Answer

Increase the PartitionOwnershipExpirationInterval to a value that is significantly greater than the LoadBalancingUpdateInterval, such as 45 seconds.
Increasing the PartitionOwnershipExpirationInterval to a value greater than the LoadBalancingUpdateInterval ensures that an instance has sufficient opportunity to renew its ownership claim before another instance considers the partition orphaned. This stabilizes partition distribution across scale-out instances and resolves the thrashing behavior.

Step-by-Step Solution

1
Analyze the client configuration and identify the relationship between the load balancing and expiration intervals.
The LoadBalancingUpdateInterval is set to 1515 seconds, while the PartitionOwnershipExpirationInterval is set to 1010 seconds.
This shows that the partition ownership claim expires before the client has a chance to execute its next periodic check to renew it.
2
Determine the impact of the interval mismatch on scale-out instances.
When Instance 1 claims a partition, it is expected to hold it for 1010 seconds. However, Instance 1 only attempts to renew ownership every 1515 seconds. Between second 1010 and second 1515, the partition appears unowned (expired) to other instances.
This allows Instance 2 or other instances to claim the partition during their own load-balancing passes, causing constant reassignment (thrashing).
3
Select the correct SDK configuration change to stabilize partition assignment.
Increase the PartitionOwnershipExpirationInterval to a value greater than the LoadBalancingUpdateInterval (e.g., 4545 seconds).
This ensures that even if a load balancing pass is slightly delayed, the partition lease remains valid, giving the owning instance sufficient time to renew it.

Key Concept

Partition ownership and load balancing configuration in EventProcessorClient
Estimated Time:2m 0s
Rate this question

Topics

Implement Azure Event Hubs Solutions