Instant Recovery with Write-Ahead Logging: Page Repair, System Restart, Media Restore, and System Failover, Second Edition

Graefe, Goetz; Guy, Wey; Sauer, Caetano

doi:10.2200/s00710ed2v01y201603dtm044

Cited by 14 publications

(10 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Efficient access to logs has been widely explored in academia, and one example is Instant Recovery [Graefe et al 2016], which proposes a structure to facilitate random access to the log, without hindering log write performance. The work describes a log structure geared towards single-page recovery, aided by two data structures, an index that finds a given page in a reduced time, and a transaction log chain, which links records related to a single transaction across pages.…”

Section: Efficient Auditingmentioning

confidence: 99%

The Impact of Privacy Regulations on DB Systems

Machado

Amora

2021

jidm

View full text Add to dashboard Cite

Personal data usage and collection are activities that used to grow unrestricted. However, several laws in the physical world ensure rights to people regarding their privacy and information usage. In the last years, legislators passed many laws, regulations, and acts to replicate these rights to the digital world. By doing so, new constraints, rights, and duties appear on every component of the data usage and collection workflow. In this paper, we discuss legislations’ implications, identifying impacts that these regulations introduce to current DBMS, and survey recent works that aim to solve the problems raised by these impacts, highlighting research opportunities and identifying how solutions can be achieved for the problems.

show abstract

Section: Efficient Auditingmentioning

confidence: 99%

The Impact of Privacy Regulations on DB Systems

Machado

Amora

2021

jidm

View full text Add to dashboard Cite

show abstract

“…Instant recovery [16] is a family of algorithms that allow new transactions to execute immediately after a system or even media failure, including localized single-page corruptions. A complementary technique is constant-time recovery [3], which exploits multi-version storage to allow instantaneous undo of arbitrarily long user transactions.…”

Section: Related Workmentioning

confidence: 99%

Rethinking Logging, Checkpoints, and Recovery for High-Performance Storage Engines

Haubenschild

Sauer

Neumann

et al. 2020

Proceedings of the 2020 ACM SIGMOD International Conference on Management of Data

Self Cite

View full text Add to dashboard Cite

For decades, ARIES has been the standard for logging and recovery in database systems. ARIES offers important features like support for arbitrary workloads, fuzzy checkpoints, and transparent index recovery. Nevertheless, many modern inmemory database systems use more lightweight approaches that have less overhead and better multi-core scalability but only work well for the in-memory setting. Recently, a new class of high-performance storage engines has emerged, which exploit fast SSDs to achieve performance close to pure in-memory systems but also allow out-of-memory workloads. For these systems, ARIES is too slow whereas inmemory logging proposals are not applicable. In this work, we propose a new logging and recovery design that supports incremental and fuzzy checkpointing, index recovery, out-of-memory workloads, and low-latency transaction commits. Our continuous checkpointing algorithm guarantees bounded recovery time. Using per-thread logging with minimal synchronization, our implementation achieves near-linear scalability on multi-core CPUs. We implemented and evaluated these techniques in our LeanStore storage engine. For working sets that fit in main memory, we achieve performance close to that of an in-memory approach, even with logging, checkpointing, and dirty page writing enabled. For the out-of-memory scenario, we outperform a state-of-the-art ARIES implementation by a factor of two.

show abstract

“…5. Then, log replay is performed to bring the segment to its most recent state, after which it can be written back into a replacement device (6).…”

Section: B Restore Algorithmmentioning

confidence: 99%

“…Instant restart [6] is an orthogonal technique that provides on-demand, incremental data access following a system failure. While the goals are similar, the design and implementation of instant restore require quite different techniques.…”

Section: A Failure Classes and Assumptionsmentioning

confidence: 99%

“…Finally, Section IV presents an empirical evaluation, while Section V concludes this paper. A high-level description of instant restore was previously published in a book chapter [6] among related instant recovery techniques. The additional contribution here is a much more detailed discussion of the design-including practical implementation aspects-as well as the first empirical evaluation of the technique with an open-source prototype.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Instant Restore After a Media Failure

Sauer

Graefe

Härder

2017

Advances in Databases and Information Systems

Self Cite

View full text Add to dashboard Cite

Media failures usually leave database systems unavailable for several hours until recovery is complete, especially in applications with large devices and high transaction volume. Previous work introduced a technique called single-pass restore, which increases restore bandwidth and thus substantially decreases time to repair. Instant restore goes further as it permits read/write access to any data on a device undergoing restoreeven data not yet restored-by restoring individual data segments on demand. Thus, the restore process is guided primarily by the needs of applications, and the observed mean time to repair is effectively reduced from several hours to a few seconds.This paper presents an implementation and evaluation of instant restore. The technique is incrementally implemented on a system starting with the traditional ARIES design for logging and recovery. Experiments show that the transaction latency perceived after a media failure can be cut down to less than a second and that the overhead imposed by the technique on normal processing is minimal. The net effect is that a few "nines" of availability are added to the system using simple and lowoverhead software techniques.

show abstract

Instant Recovery with Write-Ahead Logging: Page Repair, System Restart, Media Restore, and System Failover, Second Edition

Cited by 14 publications

References 17 publications

The Impact of Privacy Regulations on DB Systems

The Impact of Privacy Regulations on DB Systems

Rethinking Logging, Checkpoints, and Recovery for High-Performance Storage Engines

Instant Restore After a Media Failure

Contact Info

Product

Resources

About