Azure Functions
An Azure service that provides an event-driven serverless compute platform.
5,789 questions
One azure function has problems reading one partition on an eventhub
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(115.19999999999999, 28.000000000000004%, 21%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3ERE%3C/text%3E%3C/svg%3E)
Rickard Eriksson (Admin)
0
Reputation points
One of our Azure Functions is having issues all of a sudden to read from partition 7 of our Azure Event Hub, the same issue was seen about 3 months ago but the support team could not find anything so I recreated the function and Event Hub. This seems to have fixed the issue for a while but as said the issue has come up again. It is always the same partition that has the issue just as last time, just a different partition from last time. This is what I can find in the "Diagnose and solve problems" blade and "Event Hubs Trigger Issue Analysis" option:
While the other partitions all look great:
Our traffic is very predictable and i have checked that all partitions have about the same amount of messages going in to them.
Here is the host.json I am running with at the moment (removed logging parts):
{
"extensions": {
"eventHubs": {
"batchCheckpointFrequency": 5,
"maxEventBatchSize": 50,
"prefetchCount": 100
}
},
"version": "2.0"
}
It is a dotnet-isolated function running dotnet 8 with these nuget package versions:
<PackageReference Include="Microsoft.Azure.Functions.Worker" Version="2.0.0" />
<PackageReference Include="Microsoft.Azure.Functions.Worker.ApplicationInsights" Version="2.0.0" />
<PackageReference Include="Microsoft.Azure.Functions.Worker.Extensions.Http" Version="3.3.0" />
<PackageReference Include="Microsoft.Azure.Functions.Worker.Extensions.EventHubs" Version="6.3.6" />
<PackageReference Include="Microsoft.Azure.Functions.Worker.Sdk" Version="2.0.0" />
We have many azure functions also running the same versions etc but none of them are showing the same symptoms. There are also no error logs that I can find so I have no idea where to go from here.
Azure Functions
{count} votes
-
Jose Benjamin Solis Nolasco 1,601 Reputation points2025-04-25T13:59:31.19+00:00 @hello Rickard Eriksson (Admin) good morning I hope you are doing well,
Your Azure Function, running in a .NET 8 isolated worker process, is unable to read events from partition 7 of the Azure Event Hub, while other partitions function correctly. The "Diagnose and solve problems" blade and "Event Hubs Trigger Issue Analysis" indicate an issue specific to partition 7, with no processing activity (e.g., no events read, no checkpoints updated). The issue previously occurred with a different partition, was temporarily resolved by recreating the Function and Event Hub, and has now reappeared. Key details:
- Traffic: Predictable, with balanced message distribution across partitions.
- Configuration: host.json specifies batchCheckpointFrequency: 5, maxEventBatchSize: 50, prefetchCount: 100.
- Environment: .NET 8 isolated, Microsoft.Azure.Functions.Worker.Extensions.EventHubs v6.3.6, no error logs.
- Scope: Only one Function is affected, despite others using similar configurations.
This suggests a problem related to partition leasing, checkpointing, Event Hub service behavior, or an Azure-specific issue (e.g., network, identity, or service health).
Im going to give you some step to check you can do a quick check with this;
Potential Causes in an Azure Environment
- Partition Leasing Issues:
- Azure Functions uses the Event Processor Host (EPH) to manage partition leases, stored in Azure Blob Storage. A lease for partition 7 may be stuck, corrupted, or held by a stale Function instance, preventing the Function from reading events.
- The recurring issue (previously another partition, now partition 7) suggests a persistent leasing problem not fully resolved by recreation.
- Checkpointing Failures:
- Checkpoints for partition 7 may not be updating due to storage account issues (e.g., permissions, connectivity, or corruption), causing the Function to fail to progress. With batchCheckpointFrequency: 5, checkpoints are created after every 5 batches, which could amplify issues if a batch fails silently.
- Event Hub Service Issues:
- A specific partition (7) may be experiencing internal Event Hub service issues, such as replication delays or broker failures, which are not visible in logs but affect event availability. This could explain the recurrence after recreation, as the new Event Hub may have inherited similar issues.
- The Event Hub’s retention policy or partition distribution logic might cause delays for partition 7.
- Azure Network or Connectivity Issues:
- Network Security Groups (NSGs), Azure Firewall, or VNet configurations may intermittently block access to partition 7’s broker, especially if the Event Hub is in a Standard or Premium tier with multiple partitions.
- DNS resolution issues in Azure could affect partition-specific endpoints.
- Identity-Based Connection Issues:
- If using Microsoft Entra ID (Azure AD) for Event Hub access (supported by Microsoft.Azure.Functions.Worker.Extensions.EventHubs v5.x+), the Function’s managed identity may lack permissions for partition 7 or face token refresh issues.
- Function Runtime or Configuration Issues:
- The .NET 8 isolated worker process may encounter threading or concurrency issues (e.g., thread-pool starvation) when processing partition 7, especially with prefetchCount: 100 and maxEventBatchSize: 50.
- A subtle bug in the Event Hubs extension (v6.3.6) or misconfiguration in host.json could cause partition-specific failures.
- Consumer Group Issues:
- The Function’s consumer group may have multiple readers for partition 7, causing lease contention or duplicate processing, though this is less likely given the lack of error logs.
- Azure Storage Account Issues:
- The Azure Storage account used for leasing and checkpointing (specified in AzureWebJobsStorage) may have connectivity issues, throttling, or corrupted lease blobs for partition 7.our Azure Function, running in a .NET 8 isolated worker process, is unable to read events from partition 7 of the Azure Event Hub, while other partitions function correctly. The "Diagnose and solve problems" blade and "Event Hubs Trigger Issue Analysis" indicate an issue specific to partition 7, with no processing activity (e.g., no events read, no checkpoints updated). The issue previously occurred with a different partition, was temporarily resolved by recreating the Function and Event Hub, and has now reappeared. Key details:
- Traffic: Predictable, with balanced message distribution across partitions.
- Configuration: host.json specifies batchCheckpointFrequency: 5, maxEventBatchSize: 50, prefetchCount: 100.
- Environment: .NET 8 isolated, Microsoft.Azure.Functions.Worker.Extensions.EventHubs v6.3.6, no error logs.
- Scope: Only one Function is affected, despite others using similar configurations.
- Partition Leasing Issues:
- Azure Functions uses the Event Processor Host (EPH) to manage partition leases, stored in Azure Blob Storage. A lease for partition 7 may be stuck, corrupted, or held by a stale Function instance, preventing the Function from reading events.
- The recurring issue (previously another partition, now partition 7) suggests a persistent leasing problem not fully resolved by recreation.
- Checkpointing Failures:
- Checkpoints for partition 7 may not be updating due to storage account issues (e.g., permissions, connectivity, or corruption), causing the Function to fail to progress. With batchCheckpointFrequency: 5, checkpoints are created after every 5 batches, which could amplify issues if a batch fails silently.
- Event Hub Service Issues:
- A specific partition (7) may be experiencing internal Event Hub service issues, such as replication delays or broker failures, which are not visible in logs but affect event availability. This could explain the recurrence after recreation, as the new Event Hub may have inherited similar issues.
- The Event Hub’s retention policy or partition distribution logic might cause delays for partition 7.
- Azure Network or Connectivity Issues:
- Network Security Groups (NSGs), Azure Firewall, or VNet configurations may intermittently block access to partition 7’s broker, especially if the Event Hub is in a Standard or Premium tier with multiple partitions.
- DNS resolution issues in Azure could affect partition-specific endpoints.
- Identity-Based Connection Issues:
- If using Microsoft Entra ID (Azure AD) for Event Hub access (supported by Microsoft.Azure.Functions.Worker.Extensions.EventHubs v5.x+), the Function’s managed identity may lack permissions for partition 7 or face token refresh issues.
- Function Runtime or Configuration Issues:
- The .NET 8 isolated worker process may encounter threading or concurrency issues (e.g., thread-pool starvation) when processing partition 7, especially with prefetchCount: 100 and maxEventBatchSize: 50.
- A subtle bug in the Event Hubs extension (v6.3.6) or misconfiguration in host.json could cause partition-specific failures.
- Consumer Group Issues:
- The Function’s consumer group may have multiple readers for partition 7, causing lease contention or duplicate processing, though this is less likely given the lack of error logs.
- Azure Storage Account Issues:
- The Azure Storage account used for leasing and checkpointing (specified in AzureWebJobsStorage) may have connectivity issues, throttling, or corrupted lease blobs for partition 7.
😊 If my answer helped you resolve your issue, please consider marking it as the correct answer. This helps others in the community find solutions more easily. Thanks!
-
Jose Benjamin Solis Nolasco 1,601 Reputation points
2025-04-25T14:04:10.9166667+00:00 Just following up, Do you need more guidance or assistance ?
-
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(176, 40%, 27%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EKA%3C/text%3E%3C/svg%3E)
Khadeer Ali 5,065 Reputation points Microsoft External Staff Moderator2025-04-28T14:07:59.5433333+00:00 Just checking in to see if the answer provided by helped resolve your query.
-
LeelaRajeshSayana-MSFT 17,761 Reputation points Moderator2025-04-30T00:08:26.92+00:00 @Rickard Eriksson (Admin) I am following up to see if you require any additional assistance on this issue. Kindly share the details requested in the private comments to help us investigate this issue further if you still seek assistance on this.
-
Rickard Eriksson (Admin) 0 Reputation points
2025-05-05T09:07:54.6633333+00:00 Hi! Thank you for the detailed explanations! We have not been able to work out the issue yet and I am a bit unsure how to proceed, as I mentioned I cannot find any error logs at all anywhere. Neither the function, storage account nor the azure event hub seems to log any issues at all and the issue just now resurfaced after about 90 days.
Would it be of use to try and remove the checkpoint for the partition in the storage account to see if it could be the "corrupted" checkpoint issue?
It also does not seem to matter how it is deployed since last time it was in a consumption plan and now in a dedicated plan.
Is there a way for me to check if there is a problem with the eventhub or the storage account in a more detailed way to see if it is one of the other issues you mentioned?
-
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(192, 28.999999999999996%, 28%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EGK%3C/text%3E%3C/svg%3E)
Gaurav Kumar 380 Reputation points Microsoft External Staff Moderator2025-05-15T04:29:02.7166667+00:00 Hi @Rickard Eriksson (Admin) ,
Following up to check if the issue is resolved or if you need any further assistance. Let me know how it's going!
Sign in to comment
Sign in to answer