Azure serves up pre-warmed Functions to offset cold starts

Azure serves up pre-warmed Functions to offset cold starts

Microsoft has unveiled a number of enhancements for its Azure Functions serverless computing service, including PowerShell support for event-driven automation, extended language support and a way to avoid what it calls cold starts.

The string of updates were announced at Microsoft’s Ignite developer conference in Orlando, Florida, demonstrating the firm’s continued support for so-called serverless computing.

Among the announcements is general availability of the Azure Functions Premium plan, the major feature of which is support for pre-warmed instances. These are intended to avoid the situation where an event triggers a function, but if there is no instance already running there will be a delay while one is prepared – what the serverless world calls a cold start.

With the Premium plan, users can choose to have their code pre-warmed on a specified number of instances. By having a buffer of pre-warmed instances, this will effectively avoid cold start latencies. Users need to keep at least one instance running and available at all times the plan is active.

Other features of the Premium plan include a longer maximum execution time of up to 60 minutes, greater scale out, with up to 100 instances in select regions, support for Python and Linux OS hosting and enhanced regional virtual networking support.

Microsoft said that language support in Azure Functions has now been extended to include .NET Core 3 and Python 3.7.

Also generally available is PowerShell support in Azure Functions, which enables the event-driven programming model to be used for infrastructure management and scripting tasks. According to Microsoft, hundreds of built-in Azure connectors are available to automatically respond to activity not only in Azure services, but also in third party solutions and on-premises resources as well.

Serverless apps can now easily incorporate secrets management without any code changes. This enables the secure storage of sensitive information such as certificates, connection strings, or passwords by including references to Azure Key Vault secrets in the code’s application settings.

Also new is integration with Azure Monitor Logs for both Azure Functions and Azure App Service. This means that developers using Azure can now get log telemetry in a single workspace where they can analyse collected data and check on overall performance status.