Allow to configure sticky reads

### Motivation

Currently the BK client is issuing the read requests in round-robin fashion across all the bookies in the write set.

One issue with this approach is that it's not taking full advantage of the read-ahead cache, either explicit (like in `DbLedgerStorage`) or implicit (by reading data through Linux page cache which will do some prefetching).

With `e=2`, `w=2`, when we read `e-0` from `bookie-1` and `e-1` from `bookie-2`, we fail to take advantage of the fact that `bookie-1` will have already `e-1` in memory.

Effectively with `e-2`, `w-2` the disk read IO will be doubled, compared to the amount of data served to BK clients. The larger the quorum, the bigger will be overhead (eg: `e=5`, `w=5` will lead to 5x reads from disk).

### Changes

Added a BK client flag for "sticky reads". When reading from a ledger that has `E=W` (every bookie has all the entries), the sticky reads will direct all read request to 1 single bookie in the ensemble.

Reviewers: Enrico Olivelli <eolivelli@gmail.com>, Sijie Guo <sijie@apache.org>

This closes #1808 from merlimat/single-bookie-per-ledger-read-master
6 files changed
tree: d5bcac98966098ac4409b1aeeed60e3cca6632aa
  1. .github/
  2. .test-infra/
  3. .travis_scripts/
  4. bin/
  5. bookkeeper-benchmark/
  6. bookkeeper-common/
  7. bookkeeper-common-allocator/
  8. bookkeeper-dist/
  9. bookkeeper-http/
  10. bookkeeper-proto/
  11. bookkeeper-server/
  12. bookkeeper-stats/
  13. bookkeeper-stats-providers/
  14. buildtools/
  15. circe-checksum/
  16. conf/
  17. cpu-affinity/
  18. deploy/
  19. dev/
  20. docker/
  21. metadata-drivers/
  22. microbenchmarks/
  23. shaded/
  24. site/
  25. stats/
  26. stream/
  27. tests/
  28. tools/
  29. .gitignore
  30. .travis.yml
  31. LICENSE
  32. NOTICE
  33. pom.xml
  34. README.md
README.md

Build Status Build Status Coverage Status Maven Central

Apache BookKeeper

Apache BookKeeper is a scalable, fault tolerant and low latency storage service optimized for append-only workloads.

It is suitable for being used in following scenarios:

  • WAL (Write-Ahead-Logging), e.g. HDFS NameNode.
  • Message Store, e.g. Apache Pulsar.
  • Offset/Cursor Store, e.g. Apache Pulsar.
  • Object/Blob Store, e.g. storing state machine snapshots.

Get Started

  • Concepts: Start with the basic concepts of Apache BookKeeper. This will help you to fully understand the other parts of the documentation.
  • Getting Started to setup BookKeeper to write logs.

Documentation

Developers

You can also read Turning Ledgers into Logs to learn how to turn ledgers into continuous log streams. If you are looking for a high level log stream API, you can checkout DistributedLog.

Administrators

Contributors

Get In Touch

Report a Bug

For filing bugs, suggesting improvements, or requesting new features, help us out by opening a Github issue or opening an Apache jira.

Need Help?

Subscribe or mail the user@bookkeeper.apache.org list - Ask questions, find answers, and also help other users.

Subscribe or mail the dev@bookkeeper.apache.org list - Join development discussions, propose new ideas and connect with contributors.

Join us on Slack - This is the most immediate way to connect with Apache BookKeeper committers and contributors.

Contributing

We feel that a welcoming open community is important and welcome contributions.

Contributing Code

  1. See Developer Setup to get your local environment setup.

  2. Take a look at our open issues: JIRA Issues Github Issues.

  3. Review our coding style and follow our pull requests to learn about our conventions.

  4. Make your changes according to our contribution guide.

Improving Website and Documentation

  1. See Building the website and documentation on how to build the website and documentation.