Evaluating serverless computing in clouds
Applications utilization of cloud resources is expected to be completely transformed by FaaS. Nevertheless, because initializing their code and data dependencies is required before a function can start executing, it suffers from cold-start issues. The overhead associated with cold starts can be redu...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Final Year Project |
| Language: | English |
| Published: |
Nanyang Technological University
2023
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/166056 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Nanyang Technological University |
| Language: | English |
| Summary: | Applications utilization of cloud resources is expected to be completely transformed by FaaS. Nevertheless, because initializing their code and data dependencies is required before a function can start executing, it suffers from cold-start issues. The overhead associated with cold starts can be reduced by keeping functions alive and warm after they have completed execution. Because to the variety of FaaS workloads, keep-alive strategies must maintain functions based on their resource and usage characteristics, which is difficult.
Caching Algorithms can be used as Keep-Alive policies. Therefore, we explore various caching algorithms as Keep-Alive policies to study their performance. We attempt to provide a balance between cold start overhead and server utilization. In this report, we provide an experimental study of caching based algorithms and compare it to the standard TTL keep-alive policy with the help of a DES. |
|---|