Function as a Service (FaaS) is a serverless cloud computing model that enables customers to encapsulate their business logic in functions. The platform automatically executes these functions each time a specified event occurs and are terminated once the triggering event is processed. Function as a Service (FaaS) workflows are often supported by ready-to-use and fully managed services hosted by cloud providers belonging to the so-called Backend as a Service (BaaS). The integration of these two serverless models offers comprehensive support to developers, enabling them to focus on business logic development while benefiting from a fully managed and automated infrastructure. However, state-of-the-art FaaS platforms currently lack direct support for integrating Backend as a Service (BaaS) services in FaaS functions, often resulting in service calls being hard-coded within function code. This approach reduces the modularity of functions, enforces vendor lock-in, and negatively impacts the performance of FaaS workloads.In this work, we propose SFIOC, an architecture that facilitates the integration of BaaS services into FaaS functions through Dependency Injection. SFIOC enhances existing FaaS solutions by enabling dynamic and at-runtime resolution of function service dependencies. SFIOC automates dependency management, thereby improving the maintainability and modularity of serverless functions while preserving workflow performance.The capabilities of our solution are demonstrated through an extensive testbed that encompasses the enhancement with SFIOC of a public cloud provider and an open-source-based private edge environment. Published in: 2025 34th International Conference on Computer Communic
Sabbioni, A., Foschini, L. (2025). SFIOC: a Platform to Support Service Dependency Injection in Serverless Functions [10.1109/icccn65249.2025.11133723].
SFIOC: a Platform to Support Service Dependency Injection in Serverless Functions
Sabbioni, Andrea;Foschini, Luca
2025
Abstract
Function as a Service (FaaS) is a serverless cloud computing model that enables customers to encapsulate their business logic in functions. The platform automatically executes these functions each time a specified event occurs and are terminated once the triggering event is processed. Function as a Service (FaaS) workflows are often supported by ready-to-use and fully managed services hosted by cloud providers belonging to the so-called Backend as a Service (BaaS). The integration of these two serverless models offers comprehensive support to developers, enabling them to focus on business logic development while benefiting from a fully managed and automated infrastructure. However, state-of-the-art FaaS platforms currently lack direct support for integrating Backend as a Service (BaaS) services in FaaS functions, often resulting in service calls being hard-coded within function code. This approach reduces the modularity of functions, enforces vendor lock-in, and negatively impacts the performance of FaaS workloads.In this work, we propose SFIOC, an architecture that facilitates the integration of BaaS services into FaaS functions through Dependency Injection. SFIOC enhances existing FaaS solutions by enabling dynamic and at-runtime resolution of function service dependencies. SFIOC automates dependency management, thereby improving the maintainability and modularity of serverless functions while preserving workflow performance.The capabilities of our solution are demonstrated through an extensive testbed that encompasses the enhancement with SFIOC of a public cloud provider and an open-source-based private edge environment. Published in: 2025 34th International Conference on Computer Communic| File | Dimensione | Formato | |
|---|---|---|---|
|
Serverless_Dependency_Injection.pdf
accesso aperto
Tipo:
Postprint / Author's Accepted Manuscript (AAM) - versione accettata per la pubblicazione dopo la peer-review
Licenza:
Licenza per accesso libero gratuito
Dimensione
292.89 kB
Formato
Adobe PDF
|
292.89 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


