Azure / Azure/azure-functions-host
Allow customization of specialization timeout value & Improve log message
- Dominant language
- C#
- Stars
- 2k
- Forks
- 482
- Avg merge
- 2d 12h
- Merged PRs (30d)
- 38
Description
For out-of-process language workers, the specialization process (environment reload request) has a [30-second timeout](https://github.com/Azure/azure-functions-host/blob/91e0bf7eca21fb9bb380450cf9d148745b0f4870/src/WebJobs.Script/Workers/ProcessManagement/WorkerProcessCountOptions.cs#L48). If a customer's function app takes longer than 30 seconds to start, the timeout will trigger, resulting in an exception like the one below in the logs
> System.AggregateException : One or more errors occurred. (The operation has timed out.) ---> The operation has timed out. ---> (Inner Exception #0) System.TimeoutException : The operation has timed out. at Microsoft.Azure.WebJobs.Script.Grpc.GrpcWorkerChannel.PendingItem.OnTimeout() at /src/azure-functions-host/src/WebJobs.Script.Grpc/Channel/GrpcWorkerChannel.cs : 1799 at System.Runtime.ExceptionServices.ExceptionDispatchInfo.Throw() at async Microsoft.Azure.WebJobs.Script.Workers.Rpc.WebHostRpcWorkerChannelManager.SpecializeAsync() at /src/azure-functions-host/src/WebJobs.Script/Workers/Rpc/WebHostRpcWorkerChannelManager.cs : 134 at System.Runtime.ExceptionServices.ExceptionDispatchInfo.Throw() at async Microsoft.Azure.WebJobs.Script.WebHost.StandbyManager.SpecializeHostCoreAsync() at /src/azure-functions-host/src/WebJobs.Script.WebHost/Standby/StandbyManager.cs : 117<---
Currently, there is no way for customers to configure this timeout. A solution is needed to allow customization of this value. Additionally, the exception message should be improved for clarity, indicating that the specialization has failed if the app takes more than 30 seconds to start.
For now, .NET Isolated customers can work around this issue by disabling the specialization code path. This can be done by setting the `WEBSITE_USE_PLACEHOLDER_DOTNETISOLATED` app setting to `0`
Contributor guide
Research direction
Start with WorkerProcessCountOptions.cs at the linked specialization timeout, then trace the timeout and exception through GrpcWorkerChannel.cs, WebHostRpcWorkerChannelManager.cs, and StandbyManager.cs. Done means the specialization timeout can be customized and the resulting log clearly identifies specialization failure when startup exceeds the configured value.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- azure, csharp
- Domain
- backend, cloud
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Active
- Clarity
- Mostly clear
- Newbie friendliness
- 55/100