Migrating IBM Cloud Functions to Code Engine

IBM Cloud Functions has been IBM’s recommended Functions-as-a-Service solution, based on IBM’s public cloud. Available in six data centers around the world, IBM Cloud Functions serves the needs of customers since its inception.

However, the increased demands and additional requirements that our customers have been articulating, lead us to constantly innovate, improve, and evolve the IBM Functions-as-a-Service technologies and offerings. As a result and to address these demands, IBM is adding Functions-as-a-Service technology to IBM Cloud® Code Engine. For limitations when you use functions in Code Engine, see Function limitations.

Comparing Code Engine to Cloud Functions

The SOC2, ISO32k, BSI C5, and Financial Services Cloud-certified IBM Cloud Code Engine service is available in nine (9) regions. As a fully managed, serverless platform, it runs a broad variety of customer workloads as containers, batch jobs, applications, and functions. You can access Code Engine in the IBM public cloud.

The new Functions-as-a-Service capabilities enable Code Engine customers to perform short running, run-to-completion type of workloads. Whether you are a Cloud Functions customer or you are new to Functions-as-a-Service, you can learn more about running functions in IBM Cloud Code Engine. You can even take a tutorial.

Functions-as-a-Service in Code Engine offers an improved server-less value proposition in terms of user experience, usage patterns, security, and total cost of ownership.

Functions in Code Engine simplifies your user experience, with the following options.

Integrates fully with the Code Engine development flow and features.
Includes common programming languages and provides optimized managed runtimes.
Offers on-demand code execution with low cold start latency and rapid scale-out.
Offers a Web-URL based function invocation mechanism with feature-rich support for web applications.
Supported security capabilities such as access to private repositories and registries.

With IBM Cloud Code Engine functions, you can use your favorite programming language to write lightweight code that runs snippets of business logic in a scalable way. You can run code in response to HTTP requests from applications or in response to IBM Cloud services and external events.

Rather than running a container image, Code Engine functions instead, runs a code bundle. A code bundle is a collection of files that represents your function code. This code bundle is injected into the runtime container. Your code bundle is created by Code Engine and is stored in container registry or inline with the function. A code bundle is not a Open Container Initiative (OCI) standard container image.

Key capabilities

When you migrate your IBM Cloud Functions based workloads, consider the following capabilities and strategies.

Supports an evolving list of managed runtimes. For more information, see Runtimes.
- Node.js Release Version 18
- Python Release Version 3.11
Provides optimized CPU and memory combinations.

In addition, Code Engine includes Function limits, as well as project quotas.

What Code Engine entity is best for my workload?

Code Engine applications serve long running, complex, compute tasks on a scalable compute infrastructure.

Code Engine jobs are used for low-parallelism, scheduled workloads that can require high resources.

Code Engine functions can perform lightweight, short running, data transformations as results of external events. They can also serve the content for dynamic, data elements on Web pages. In general, functions can perform scalable, short running parallel tasks that must be finished in a defined or short time.

Code Engine functions offers a straight-forward programming model that uses source code snippets of supported programming languages. These code snippets are used “inline” in a function definition with no need to compile your code first.

Comparing Code Engine apps, jobs, and functions
Characteristic	App	Job	Function
Execution time (duration)	Long-running (10 minutes per request)	Long-running (up to 24 hours)	Short-running (2 minutes or less)
Startup latency	Medium	Scheduled start	Low
Termination	Run-continuously	Run-to-completion	Run-to-completion
Invocation	On request or permanently running	Scheduled	On request, instant
Programming Model	Container-based build and execution	Container-based build and execution	Language-specific source code files and dependency metadata
Parallelism	Parallel execution, flexible	Low to medium parallel execution	High parallel execution
Scale-out	Based on number of requests	Based on job workload definition	Based on events or direct invocations
Optimized for	Long running, highly complex workload and on-demand scale-out	Scheduled or planned workloads with high resource demands	Startup time and rapid scale-out

Migrating IBM Cloud Functions Actions to Code Engine Functions FAQ

How can I process a bulk-load of computations?

If you process a bulk-load of computations that require high CPU and memory resources and must be finished in less than n hours, you can migrate your program logic into a Code Engine job and then schedule your jobs to run daily. For more information, see Working with jobs and job runs.

I used Cloud Function to include dynamic elements for my web application. Can I move to Code Engine Functions?

You can convert your action to a Code Engine function and then use the provided function URL to invoke and return the required dynamic content. See Working with Functions to get started.

Can I trigger my function code?

Yes. Code Engine supports cron jobs (alarms) for functions. Other types of event producers (or triggers) are only supported for applications and jobs. You can create your code as an application or job, and then add a subscription to an event producer, including cron jobs, Object storage, and Event Streams (Kafka) data. For more information, see Getting started with subscriptions.

Can my function be accessed through a public URL?

A Code Engine function includes a public URL, that is provided for you when you create it. You can also use a custom domain for your function.

How can I secure my functions?

While your code can be secured in a Git repository or container registry, because your function URL is public, authorization to invoke your function must be implemented by the function code itself.

Can I include dynamic elements?

You can include dynamic elements that are supported by Code Engine. For example, if you are using a Cloudant database and invoke Cloud Functions actions based on data changes, then you cannot migrate as Code Engine does not support subscribing to a Cloudant database. You can, however, subscribe to IBM Cloud Object Storage.

Can I use sequences to chain my functions together?

Code Engine does not include support for sequences. However, because any function can be invoked by calling its private or public endpoint, you can chain your functions together with a series of REST calls. This logic must be added directly to your function code.

Can I use `whisk.system` actions?

If you are using Cloud Functions that are based on whisk.system actions, you can export your whisk.system action with the ibmcloud fn action get command and including the --save option. The output contains the source code and references to required libraries. Then, create a Code Engine Function, based on the exported source artifacts. While a certain level of artifact compatibility is retained, you cannot export the Action metadata directly into Code Engine Functions.

Can I bind my function to service credentials?

Yes, service binds are supported! See Working with service bindings to integrate IBM Cloud services with Code Engine.

Where can I find information about my in progress and finished Code Engine Function runs?

If you created your function from source code, you can view build logs.

Otherwise, you can use set up IBM Cloud Logs to view Platform logs, which contain information about your function invocation (meta information) as well as log messages emitted by the function code.

Does Code Engine provide an OpenAPI specification for the deployed function?

No, Code Engine does not generate or provide an OpenAPI specification for the functions you deploy. There are packages and tools available for many programming languages to generate an OpenAPI specification from code.