c-Lock: Dynamic Lock-coalescing for Latency-sensitive Distributed Locking

نویسندگان

  • Adnan Hassan
  • Naresh Rapolu
  • Ananth Grama
  • Wojciech Szpankowski
چکیده

Transactional updates to groups of keys are common to a wide class of distributed systems, spanning MapReduce-based workloads, distributed key-value stores and conventional distributed object databases. Such updates are typically implemented using nested transactions/ locks. The cost of such updates is critically dependent on the nesting depth, the contention on any one lock, and the coherence of key-groups. In this paper, we analyze techniques for coalescing key updates and associated locks to reduce nesting depth. Specifically, we demonstrate that in real workloads, the coherence of keygroup updates enables us to effectively coalesce locks to gain significant performance improvements. We prove our results through rigorous average case analysis, and support it using detailed simulations.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Conflict Control Locking in Distributed Cooperative Graphics Editors

The communication latency presents a major challenge to achieving high responsiveness for Internet-based cooperative editing systems. In this paper, we propose conflict control locking (post-locking) schemes for conflict resolution in real-time object-based cooperative graphics editors. With these schemes, no locking request is needed before applying an editing operation to an object, instead, ...

متن کامل

Efficient Locking and Caching of Data in the Multisystem Shard Disks Transaction Environment

This paper describes a technique for use when multiple instances of a data base management system (DBMS), each with its own cache (buffer pool), can directly read and modify any data stored on a set of shared disks. Global locking and coherency control protocols are necessary in this context for assuring transaction consistency and for maintaining coherency of the data cached in the multiple ca...

متن کامل

Operating System Design and Implementation 1994. Experiences with Locking in a NUMA Multiprocessor Operating System Kernel

We describe the locking architecture of a new operating system, HURRICANE, designed for large scale shared-memory multiprocessors. Many papers already describe kernel locking techniques, and some of the techniques we use have been previously described by others. However, our work is novel in the particular combination of techniques used, as well as several of the individual techniques themselve...

متن کامل

Permit Based Locking

The Permit based Locking design pattern provides fast acquisition and release of locks without network communication even in distributed systems build around a central lock server. Instead of plain locks, the server manages "permits to lock", and passes them to those clients which probably need the lock next. If a client hosts a permit, the associated lock can be acquired immediately without se...

متن کامل

Design Characteristics and Dynamic Modeling of a Cooperative Dual-Arm- Lock Manipulator

Recent developments in the area of smart structures indicate that variable geometry / stiffness truss network is of fundamental importance in designing smart transformable structures and systems for space applications. This paper presents the conceptual design and dynamic modeling of a cooperative re-configurabel dual-arm robotic structure called Dual-Arm Cam-Lock Manipulator. The Manipulator i...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2011