7/2/2015EECS 584, Fall Bigtable: A Distributed Storage System for Structured Data Jing Zhang Reference: Handling Large Datasets at Google: Current System and Future Directions, Jeff Dean

7/2/2015 EECS 584, Fall Outline Motivation Data Model APIs Building Blocks Implementation Refinement Evaluation

7/2/2015 EECS 584, Fall Outline Motivation Data Model APIs Building Blocks Implementation Refinement Evaluation

Google’s Motivation – Scale! Scale Problem –Lots of data –Millions of machines –Different project/applications –Hundreds of millions of users Storage for (semi-)structured data No commercial system big enough –Couldn’t afford if there was one Low-level storage optimization help performance significantly – Much harder to do when running on top of a database layer 7/2/2015 EECS 584, Fall 20114

Bigtable Distributed multi-level map Fault-tolerant, persistent Scalable –Thousands of servers –Terabytes of in-memory data –Petabyte of disk-based data –Millions of reads/writes per second, efficient scans Self-managing –Servers can be added/removed dynamically –Servers adjust to load imbalance 7/2/2015 EECS 584, Fall 20115

Real Applications 7/2/2015 EECS 584, Fall 20116

7/2/2015 EECS 584, Fall Outline Motivation Data Model APIs Building Blocks Implementation Refinement Evaluation

Data Model a sparse, distributed persistent multi- dimensional sorted map (row, column, timestamp) -> cell contents 7/2/2015 EECS 584, Fall 20118

Data Model Rows –Arbitrary string –Access to data in a row is atomic –Ordered lexicographically 7/2/2015 EECS 584, Fall 20119

Data Model Column –Tow-level name structure: family: qualifier –Column Family is the unit of access control 7/2/2015 EECS 584, Fall

Data Model Timestamps –Store different versions of data in a cell –Lookup options Return most recent K values Return all values 7/2/2015 EECS 584, Fall

Data Model The row range for a table is dynamically partitioned Each row range is called a tablet Tablet is the unit for distribution and load balancing 7/2/2015 EECS 584, Fall

7/2/2015 EECS 584, Fall Outline Motivation Data Model APIs Building Blocks Implementation Refinement Evaluation

APIs Metadata operations –Create/delete tables, column families, change metadata Writes –Set(): write cells in a row –DeleteCells(): delete cells in a row –DeleteRow(): delete all cells in a row Reads –Scanner: read arbitrary cells in a bigtable Each row read is atomic Can restrict returned rows to a particular range Can ask for just data from 1 row, all rows, etc. Can ask for all columns, just certain column families, or specific columns 7/2/2015 EECS 584, Fall

7/2/2015 EECS 584, Fall Outline Motivation Data Model APIs Building Blocks Implementation Refinement Evaluation

Typical Cluster 7/2/2015 EECS 584, Fall Shared pool of machines that also run other distributed applications

Building Blocks Google File System (GFS) –stores persistent data (SSTable file format) Scheduler –schedules jobs onto machines Chubby –Lock service: distributed lock manager –master election, location bootstrapping MapReduce (optional) –Data processing –Read/write Bigtable data 7/2/2015 EECS 584, Fall

Chubby {lock/file/name} service Coarse-grained locks Each clients has a session with Chubby. –The session expires if it is unable to renew its session lease within the lease expiration time. 5 replicas, need a majority vote to be active Also an OSDI ’06 Paper 7/2/2015 EECS 584, Fall

7/2/2015 EECS 584, Fall Outline Motivation Overall Architecture & Building Blocks Data Model APIs Implementation Refinement Evaluation

Implementation Single-master distributed system Three major components –Library that linked into every client –One master server Assigning tablets to tablet servers Detecting addition and expiration of tablet servers Balancing tablet-server load Garbage collection Metadata Operations –Many tablet servers Tablet servers handle read and write requests to its table Splits tablets that have grown too large 7/2/2015 EECS 584, Fall

Implementation 7/2/2015 EECS 584, Fall

Tablets Each Tablets is assigned to one tablet server. –Tablet holds contiguous range of rows Clients can often choose row keys to achieve locality –Aim for ~100MB to 200MB of data per tablet Tablet server is responsible for ~100 tablets –Fast recovery: 100 machines each pick up 1 tablet for failed machine –Fine-grained load balancing: Migrate tablets away from overloaded machine Master makes load-balancing decisions 7/2/2015 EECS 584, Fall

How to locate a Tablet? Given a row, how do clients find the location of the tablet whose row range covers the target row? 7/2/2015 EECS 584, Fall METADATA: Key: table id + end row, Data: location Aggressive Caching and Prefetching at Client side

Tablet Assignment Each tablet is assigned to one tablet server at a time. Master server keeps track of the set of live tablet servers and current assignments of tablets to servers. When a tablet is unassigned, master assigns the tablet to an tablet server with sufficient room. It uses Chubby to monitor health of tablet servers, and restart/replace failed servers. 7/2/2015 EECS 584, Fall

Tablet Assignment Chubby –Tablet server registers itself by getting a lock in a specific directory chubby Chubby gives “lease” on lock, must be renewed periodically Server loses lock if it gets disconnected –Master monitors this directory to find which servers exist/are alive If server not contactable/has lost lock, master grabs lock and reassigns tablets GFS replicates data. Prefer to start tablet server on same machine that the data is already at 7/2/2015 EECS 584, Fall

7/2/2015 EECS 584, Fall Outline Motivation Overall Architecture & Building Blocks Data Model APIs Implementation Refinement Evaluation

Refinement – Locality groups & Compression Locality Groups –Can group multiple column families into a locality group Separate SSTable is created for each locality group in each tablet. –Segregating columns families that are not typically accessed together enables more efficient reads. In WebTable, page metadata can be in one group and contents of the page in another group. Compression –Many opportunities for compression Similar values in the cell at different timestamps Similar values in different columns Similar values across adjacent rows 7/2/2015 EECS 584, Fall

7/2/2015 EECS 584, Fall Outline Motivation Overall Architecture & Building Blocks Data Model APIs Implementation Refinement Evaluation

Performance - Scaling 7/2/2015 EECS 584, Fall As the number of tablet servers is increased by a factor of 500: –Performance of random reads from memory increases by a factor of 300. –Performance of scans increases by a factor of 260. Not Linear! WHY?

Not linearly? Load Imbalance –Competitions with other processes Network CPU –Rebalancing algorithm does not work perfectly Reduce the number of tablet movement Load shifted around as the benchmark progresses 7/2/2015 EECS 584, Fall