Mitigating Cold Start Problem in Serverless Computing with Function Fusion
As Artificial Intelligence (AI) is becoming ubiquitous in many applications, serverless computing is also emerging as a building block for developing cloud-based AI services. Serverless computing has received much interest because of its simplicity, scalability, and resource efficiency. However, due...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2021-12-01
|
Series: | Sensors |
Subjects: | |
Online Access: | https://www.mdpi.com/1424-8220/21/24/8416 |
_version_ | 1797500788418805760 |
---|---|
author | Seungjun Lee Daegun Yoon Sangho Yeo Sangyoon Oh |
author_facet | Seungjun Lee Daegun Yoon Sangho Yeo Sangyoon Oh |
author_sort | Seungjun Lee |
collection | DOAJ |
description | As Artificial Intelligence (AI) is becoming ubiquitous in many applications, serverless computing is also emerging as a building block for developing cloud-based AI services. Serverless computing has received much interest because of its simplicity, scalability, and resource efficiency. However, due to the trade-off with resource efficiency, serverless computing suffers from the cold start problem, that is, a latency between a request arrival and function execution. The cold start problem significantly influences the overall response time of workflow that consists of functions because the cold start may occur in every function within the workflow. Function fusion can be one of the solutions to mitigate the cold start latency of a workflow. If two functions are fused into a single function, the cold start of the second function is removed; however, if parallel functions are fused, the workflow response time can be increased because the parallel functions run sequentially even if the cold start latency is reduced. This study presents an approach to mitigate the cold start latency of a workflow using function fusion while considering a parallel run. First, we identify three latencies that affect response time, present a workflow response time model considering the latency, and efficiently find a fusion solution that can optimize the response time on the cold start. Our method shows a response time of 28–86% of the response time of the original workflow in five workflows. |
first_indexed | 2024-03-10T03:08:52Z |
format | Article |
id | doaj.art-99e787dc0292452cb692f8bc1ddba69b |
institution | Directory Open Access Journal |
issn | 1424-8220 |
language | English |
last_indexed | 2024-03-10T03:08:52Z |
publishDate | 2021-12-01 |
publisher | MDPI AG |
record_format | Article |
series | Sensors |
spelling | doaj.art-99e787dc0292452cb692f8bc1ddba69b2023-11-23T10:31:09ZengMDPI AGSensors1424-82202021-12-012124841610.3390/s21248416Mitigating Cold Start Problem in Serverless Computing with Function FusionSeungjun Lee0Daegun Yoon1Sangho Yeo2Sangyoon Oh3Department of Artificial Intelligence, Ajou University, Suwon 16499, KoreaDepartment of Artificial Intelligence, Ajou University, Suwon 16499, KoreaDepartment of Artificial Intelligence, Ajou University, Suwon 16499, KoreaDepartment of Artificial Intelligence, Ajou University, Suwon 16499, KoreaAs Artificial Intelligence (AI) is becoming ubiquitous in many applications, serverless computing is also emerging as a building block for developing cloud-based AI services. Serverless computing has received much interest because of its simplicity, scalability, and resource efficiency. However, due to the trade-off with resource efficiency, serverless computing suffers from the cold start problem, that is, a latency between a request arrival and function execution. The cold start problem significantly influences the overall response time of workflow that consists of functions because the cold start may occur in every function within the workflow. Function fusion can be one of the solutions to mitigate the cold start latency of a workflow. If two functions are fused into a single function, the cold start of the second function is removed; however, if parallel functions are fused, the workflow response time can be increased because the parallel functions run sequentially even if the cold start latency is reduced. This study presents an approach to mitigate the cold start latency of a workflow using function fusion while considering a parallel run. First, we identify three latencies that affect response time, present a workflow response time model considering the latency, and efficiently find a fusion solution that can optimize the response time on the cold start. Our method shows a response time of 28–86% of the response time of the original workflow in five workflows.https://www.mdpi.com/1424-8220/21/24/8416serverless computingfunction fusionserverless workflow |
spellingShingle | Seungjun Lee Daegun Yoon Sangho Yeo Sangyoon Oh Mitigating Cold Start Problem in Serverless Computing with Function Fusion Sensors serverless computing function fusion serverless workflow |
title | Mitigating Cold Start Problem in Serverless Computing with Function Fusion |
title_full | Mitigating Cold Start Problem in Serverless Computing with Function Fusion |
title_fullStr | Mitigating Cold Start Problem in Serverless Computing with Function Fusion |
title_full_unstemmed | Mitigating Cold Start Problem in Serverless Computing with Function Fusion |
title_short | Mitigating Cold Start Problem in Serverless Computing with Function Fusion |
title_sort | mitigating cold start problem in serverless computing with function fusion |
topic | serverless computing function fusion serverless workflow |
url | https://www.mdpi.com/1424-8220/21/24/8416 |
work_keys_str_mv | AT seungjunlee mitigatingcoldstartprobleminserverlesscomputingwithfunctionfusion AT daegunyoon mitigatingcoldstartprobleminserverlesscomputingwithfunctionfusion AT sanghoyeo mitigatingcoldstartprobleminserverlesscomputingwithfunctionfusion AT sangyoonoh mitigatingcoldstartprobleminserverlesscomputingwithfunctionfusion |