Towards application recoverability atop cloud-native storage
Luo, Wenqing
This item is only available for download by members of the University of Illinois community. Students, faculty, and staff at the U of I may log in with your NetID and password to view the item. If you are trying to access an Illinois-restricted dissertation or thesis, you can request a copy through your library's Inter-Library Loan office or purchase a copy directly from ProQuest.
Permalink
https://hdl.handle.net/2142/120413
Description
Title
Towards application recoverability atop cloud-native storage
Author(s)
Luo, Wenqing
Issue Date
2023-04-26
Director of Research (if dissertation) or Advisor (if thesis)
Xu, Tianyin
Ganesan, Aishwarya
Alagappan, Ramnatthan
Department of Study
Computer Science
Discipline
Computer Science
Degree Granting Institution
University of Illinois at Urbana-Champaign
Degree Name
M.S.
Degree Level
Thesis
Keyword(s)
Cloud system
Disaggregated storage
Recoverability
Abstract
Cloud-native applications, designed specifically for cloud environments, rely heavily on disaggregated storage services to manage their persistent state. These storage services enable applications to recover their state after failures, ensuring data durability and application availability. However, this thesis poses a critical question: can cloud-native applications consistently and correctly recover after experiencing failures? To investigate this, we have conducted preliminary research focusing on applications built on top of disaggregated file services.
Our main finding reveals that storage services demonstrate various post-failure behaviors, leading to unexpected states after application failures. These unpredictable states can cause severe consequences, such as data loss and application unavailability. We have identified that subtle interactions between the application and the storage service can significantly impact an application’s recoverability.
Given the importance of recoverability in cloud-native applications, we aim to bring attention to this issue and outline the steps and vision to address it. In this thesis, we delve into understanding the interactions between applications and storage services, identifying post-failure behaviors exhibited by storage services, and uncovering application vulnerabilities based on the post-failure behaviors examined. Additionally, we developed an automated tool to emulate post-failure behaviors in application persistence protocols and detect recoverability issues.
Upon examining four applications atop two disaggregated file services, we identified vulnerabilities in all applications, resulting from the post-failure behaviors of the storage services. Our automated tool successfully reproduces all discovered vulnerabilities, and we aim to enhance the tool and apply it to more applications to uncover further vulnerabilities.
Ultimately, our research seeks to contribute to the growing body of knowledge on cloudnative applications and their recoverability, ensuring the development of more reliable, resilient, and fault-tolerant systems within the cloud.
Use this login method if you
don't
have an
@illinois.edu
email address.
(Oops, I do have one)
IDEALS migrated to a new platform on June 23, 2022. If you created
your account prior to this date, you will have to reset your password
using the forgot-password link below.