Enhancing the reliability of a system executing real-time jobs is, in many cases, one of the most important design goals. A dynamically reconfigurable system offers an approach for improvement of reliability. To achieve high reliability, the most suitable recovery action must be used when a fault occurs, which means that some kind of optimal recovery strategy should be followed. This is called a dynamic recovery strategy. To satisfy the service requirements of real-time jobs with hard deadlines, a more powerful system, intuitively, should always be preferred. On the other hand, higher processing capacity means more processing modules and electronics parts, which may result in more frequent faults and a higher risk that the system will fail to complete the real-time jobs prior to their deadline. In this paper, we investigate the reliability enhancement of a real-time distributed computing system with hard deadlines through the employment of dynamic recovery strategies. Since the classical...

Due to little interference with other database activities, fuzzy checkpointing has been shown to be quite suitable for Main Memory Databases. This paper presents a new fuzzy checkpointing approach called Segmented Fuzzy Checkpointing (SFC), which can greatly reduce the log processing time during restart without or only slightly degrading the system performance. SFC uses a moving window to locate a more efficient redo-point than other fuzzy checkpointing approaches. The concept of SFC is simple and flexible. Our analytical result shows that SFC speeds up the restart operation with little overhead.

In this paper, we propose a checkpoint technique for real-time main memory database (MMDB) systems which aims at not only reducing the system recovery time, but also increasing the percent of transactions meeting their deadlines and enabling as many temporal data as possible to be used before losing their validity. The main idea of this approach is to divide the MMDB into partitions based on data types (persistent vs. temporal), update frequencies and temporal valid intervals, and checkpoint each partition independently based on its update frequency and temporal valid interval. Our simulation results show that the proposed scheme offers a significant performance improvement over the conventional fuzzy checkpoint approach, and the key factors in the design of a real-time MMDB checkpoint technique include how the MMDB is partitioned, whether temporal data of short valid intervals are checkpointed and how the checkpoints among partitions are scheduled. 1. Introduction A real-time databa...

While flash memory has been widely adopted for storage systems for various embedded systems, issues on performance and reliability have started receiving growing attention in recent years. How to provide efficient roll back and quick mounting for flash-memory file systems has become important research topics in recent years, in addition to the work on effective garbage collection and superb run-time performance. Such an observation motivates our work on the investigation of efficient initialization and crash recovery of flash-memory file systems based on log structures. A methodology is proposed for the acceleration of mounting and crash recovery for log-based file systems. A system prototype based on a well-known flash-memory file system YAFFS was implemented with performance evaluation. The experimental results show that the proposed methodology can reduce the mounting time significantly, regardless of whether the file system is properly unmounted. 1.

Checkpoint of an in-memory database is the main source of a persistent database image surviving a software crash, or a power outage, and is, together with transactions logs, a foundation for transaction durability. Since checkpoints are created simultaneously with transaction processing, they tend to decrease database throughput and increase its memory footprint. Of the current methods, most efficient are the fuzzy checkpoint algorithms that write dirty pages to disk and require transaction logs for reconstructing a consistent state. Known consistency-preserving methods suffer from excessive memory usage or a transaction-blocking behavior. In this paper, we present a consistency-preserving and memory-efficient checkpoint method. It is based on tuple shadowing as opposed to known page shadowing methods, and rearranging of tuples between pages for minimal memory usage overhead. The method’s algorithms are introduced and both analytical and experimental analysis of the proposed algorithms show significant reduction in the memory usage overhead, and up to 30 % higher transaction throughput compared with a fuzzy checkpoint method with undo/redo log. 1

Massively multiplayer online games (MMOs) have emerged as an exciting new class of applications for database technology. MMOs simulate long-lived, interactive virtual worlds, which proceed by applying updates in frames or ticks, typically at 30 or 60 Hz. In order to sustain the resulting high update rates of such games, game state is kept entirely in main memory by the game servers. Nevertheless, durability in MMOs is usually achieved by a standard DBMS implementing ARIES-style recovery. This architecture limits scalability, forcing MMO developers to either invest in high-end hardware or to over-partition their virtual worlds. In this paper, we evaluate the applicability of existing checkpoint recovery techniques developed for main-memory DBMS to MMO workloads. Our thorough experimental evaluation uses a detailed simulation model fed with update traces generated synthetically and from a prototype game server. Based on our results, we recommend MMO developers to adopt a copy-on-update scheme with a double-backup disk organization to checkpoint game state. This scheme outperforms alternatives in terms of the latency introduced in the game as well the time necessary to recover after a crash. 1.

Many applications, such as telecommunication, process control, and virtual reality, require real-time access to database. Main-memory DBMS, which becomes feasible with the increasing availability of large and relatively cheap memory, can provide better performance than disk-based systems for real-time applications. This paper presents an overall architecture of M 2 RT, a Main-Memory Real-Time DBMS, and an object-oriented design of its storage system called M 2 RTSS. M 2 RTSS provides classes that implement the core functionality of storage management, real-time transaction scheduling, and recovery. Implementation-specific information is encapsulated in these classes and extensions can be made by inheritance. With object-oriented features, M 2 RTSS can easily incorporate new development in application requirements and the result of ongoing research in realtime systems. Keywords: object-oriented design and implementation, extensibility, main-memory DBMS, real-time DBMS 1 Introdu...

In this paper we present several new ideas about concurrent shadow paging as a crash recovery method in databases. We show how to use shadow paging in a multi-user environment and describe several optimizations and ideas which significantly improve the performance and general usability of shadow paging, making it very competitive to other methods in crash recovery. Shadow paging with our optimizations appears to be much faster than log-based solutions, and it would seem to be very generally applicable. We also present some new, potentially very useful ideas which can be efficiently implemented with shadow paging but not with logging. 1 Introduction The idea of shadow paging is to never overwrite valid data. A page table is used to map logical page numbers to physical page numbers, and modified pages are always written to unused parts of the database. The page table is updated atomically at transaction commit to reflect the new state of the database. This allows arbitrary atomic transa...

This paper presents the architecture of Xmas, an extensible mainmemory storage system for high-performance and real-time database applications. Xmas not only provides the core functionality of a DBMS, such as data persistence, concurrency control, and crash recovery, but also has an extensible architecture to deal with requirements from various applications. One crucial aspect of such extensibility is that an application developer can compose applicationspecific, high-level operations with a basic set of operations provided by the system. Called composite actions in Xmas, these operations are processed by a customized Xmas server with minimum interaction with application processes, thus increasing the overall performance. 1 Introduction Many applications such as mobile communication systems, factory automation, and distributed virtual reality networks, require high-performance and real-time access to database[KGM95]. Disk-based DBMSs, however, are inadequate for these types of applica...

This paper presents a simple and effective method to reduce the size of log data for recovery in main memory databases. Fuzzy checkpointing is known to be very efficient in main memory databases due to asynchronous backup activities. By this feature, most recovery works in the past have used only physical logging schemes. Since the size of physical log records is quite large, physical logging schemes cause the recovery time to be much longer than that using logical logging schemes. In this paper, we propose a hybrid logging method that can accommodate logical logging under fuzzy checkpointing. This method significantly decreases the size of log data, and hence speeds up the recovery time. We also propose a reapplying rule in segmented MMDB, which reduces the number of log records for recovery. We evaluate the performance of the proposed methods through analytic analyses. The results show that we can reduce the size of log data to more than half, compared with those that use only physical logging.