Apache con2016final
8 likes1,881 views
The document discusses Salesforce's requirements for a distributed storage solution and their evaluation of Apache Bookkeeper as a candidate, highlighting its features like low latency, high availability, and consistency. It outlines Bookkeeper's architecture, major components, operational guarantees, recovery processes, and the community involvement surrounding the project. Additionally, enhancements and future directions for the platform are mentioned, along with its successful implementations in various companies.
Apache con2016final
- 1. Apache BookKeeper DISTRIBUTED STORE a Salesforce Use Case Venkateswararao Jujjuri (JV) Cloud Storage Architect vjujjuri@salesforce.com jujjuri@gmail.com @jvjujjuri | Twitter https://www.linkedin.com/in/jvjujjuri
- 2. Agenda Salesforce needs and requirements Hunt and Selection BookKeeper Introduction Improvements and Enhancements As Service at Scale @ Salesforce Performance Community Q & A
- 3. Salesforce Application Storage Needs Store for Persistent WAL, data, and objects Low, constant write latencies • Transaction Log, Smaller writes Low, constant Random Read latencies Highly available Append Only entries • Objects Highly Consistent for immutable data Long Term Storage Distributed and linearly scalable. On commodity hardware Low Operating Cost
- 4. What Did we consider? Build vs. Buy • Time-To-Market, resources, cost. Finalists • Ceph • A CP System • w/Unreliable reads, Read path can behave like an AP system. • Lot of effort to make it AP behavior on write path • Remember: Immutable data. • BookKeeper • CAP system, because of immutable/append only data. • Came close to what we want • Almost there but not everything.
- 5. Apache Bookkeeper A highly consistent, available, replicated, distributed log service. Immutable , append only store. Thick Client, Simple and Elegant placement policy • No Central Master • No complicated hashing/computing for placement Low latency, both on writes and reads. Runs on commodity hardware. Built for WAL use-case, but can be expanded to broader storage needs Uses ZooKeeper as consensuses service, and metadata store. Awesome Community.
- 7. Apache BookKeeper A system to reliably log streams of records. Is designed to store write ahead logs for database like applications. Inspired by and designed to solve HDFS NameNode availability deficiencies. Opensource Chronology • 2008 Open Sourced contribution to ZooKeeper • 2011 Sub-Project of ZooKeeper. • 2012 Production
- 8. Terminology Journal: Write ahead log Ledger: Log Stream Entry: Each entry of log stream Client: Library, with the application. Bookie: Server Ensemble: Set of Bookies across which a ledger is striped. Cluster: All bookies belong to a given instance of Bookkeeper Write Quorum Size: Number of replicas. Ack Quorum Size: Number of responses needed before client’s write is satisfied. LAC: Last Add Confirmed.
- 9. Major Components • Thick Client; Carries heavy weight in the protocol. • Thin Server, Bookie. Bookies never initiate any interaction with ZooKeeper or fellow Bookies. • Zookeeper monitors Bookies. • Metadata is stored on Zookeeper. • Auditor to monitor bookies and identify under replicated ledgers. • Replication workers to replicate under replicated ledger copies. Major Components
- 10. Create Ledger • Gets Writer Ledger Handle Add an entry to the Ledger • Write To the Ledger Open Ledger • Gives ReadOnly Ledger Handle. • May ask for non-recovery read handle. Get an entry from the ledger • Read from the ledger Close ledger Delete Ledger Basic Operations
- 11. Salesforce Application with BookKeeper Application Store Interface With Bookkeeper client User Library Bookies ZooKeeper Server Machine
- 12. Guarantees • If an entry has been acknowledged, it must be readable. • If an entry is read once, it must always be readable. • If write of entryID ‘n’ is successful, all entries until ‘n’ are successfully committed. Consistencies • Last Add Confirmed is consistency among readers • Fence is consistency among writers. Commitment
- 13. Out-of-order write and In-Order Ack. • Application has liberty to pre-allocate entryIDs • Multiple application threads can write in parallel. User defined Ledger Names • Not restricted by BK generated ledger Names Explicit LAC updates • Added ReadLac, WriteLac to the protocol. • Maintain both piggy-back LAC and explicit LAC simultaneously. Enhancements - In the internal branch working to push upstream
- 14. Conventional Name Space. • User defined Names • Treat LedgerId as an i-node in a file system. Disk scrubbers and Repairs • Actively hunt and repair bit-rots and corruptions Scalable Metadata Store • Separate and dedicated metadata store • Not restricted by ZK limitations Enhancements - Future
- 15. Out of order write and in order Ack 0 1 2 3 4 5 App A ( Writer ) 6 App B ( Writer ) 8 App C ( Writer ) 7
- 16. Last Add Confirmed 0 1 2 3 4 5 App A ( Writer ) 6 App B ( Writer ) 8 App C ( Writer ) 7 LAC LAC App D (Reader) X LAC
- 17. Things Do Break
- 18. What Can Happen? Client • Client Restarts • Client loses connection with zookeeper • Client loses connection with bookies. Bookie • Bookie Goes down • Disk(s) on bookie go bad, IO issues • Bookie gets disconnected from network. Zookeeper • Gets disconnected from rest of the cluster
- 19. Writing Client Crash bookie bookie bookie zookeeper What is the last entry? • Nothing happens until a reader attempts to read. • Recovery process gets initiated when a process opens the ledger for reading. • Close the ledger on zoo keeper • Identify Last entry of the ledger. • Update metadata on zookeeper with Last Add Confirmed. (LAC)
- 20. Client gets disconnected with Bookies. Either bookie is down or network between client and bookie have issues. Contact zoo keeper to get the list of available bookies. Update ensemble set, register with zookeeper. Continue with new set.
- 21. Client gets disconnected with Zookeeper. Tries to reestablish the connection. Can continue to read and write to the ledger. Until that time, no metadata operations can be performed. • Can not create a ledger • Can not open a ledger • Can not close a ledger
- 22. Reader Opens while writer is active. Application control BK guarantees correctness. Reader initiates recovery process. • Fences bookie on the zookeeper. • Informs all bookies in ensemble recovery started. • After these steps writer will get write errors.(if actively writing) • Reader contacts all bookies to learn last entry. • Replicates last entry if it doesn’t have enough replicas. • Updates zookeeper with LAC, and closes the ledger.
- 23. Recovery begins when the ledger is opened by the reader in recovery mode • Check if the ledger needs recovery (not closed) • Fence the ledger first and initiate recovery • Step1: Flag that the ledger is in recovery by update ZooKeeper state. • Step2 : Fence Bookies • Step3 : Recover the Ledger Fencing and Recovery
- 24. Ledger Fencing BookKeeper Distributed Store Ledger Write Non Recovery Read Recovery ReadFence & Recover Attempt to write
- 25. ZooKeeper Cluster B Auto Recovery Components Bookie-1 Bookie-2 Bookie-N BookKeeper Cluster Auditor (Lead) Replicator Worker Auditor (Follower) Replicator Worker Auditor (Follower) Replicator Worker Machine-1 Machine-2 Machine-N
- 26. Auditor • Starts on every Bookie machine, leader gets elected through ZooKeeper. • One active auditor per cluster. • Watch Bookie failures and manage under replicated ledgers list. Replication Workers • Responsible for performing replication to maintain quorum copies. • Can run on any machine in the cluster, usually runs on each Bookie machine. • Work on under replicated ledgers list published by the Auditor. • Pick one ledger at a time, create a lock on ZooKeeper and replicate to local bookie. • If local bookie is part of the ensemble, drop the lock and move to next one in the list. Bookie Crashes - Auto Recovery
- 27. Heterogeneous Stores and Tiered Architecture Log Store Data Store Archival Store
- 28. Clusters of storage serving App Instances Log Store Data Store Archival Store App Instance App Instance App Instance App Instance App Instance App Instance App Instance App Instance
- 29. Performance
- 30. Performance
- 31. Performance
- 32. Community Update Projects built on BookKeeper • Twitter Distributed Log : Manhattan, Pub/Sub, DeferredRPC • Yahoo Cloud Messaging • Salesforce Distributed Store. • Huawei – HDFS NameNode • HubSpot – WAL • Majordodo – Distributed Resource Manager Community • 6 PMC members • 8 Committers • 20-25 active members • 5 Enterprises actively using/contributing More Info https://cwiki.apache.org/confluence/display/BOOKKEEPER/BookKeeper+papers+and+presentations