AresDB

AresDB is a GPU-based real-time analytics storage and query engine with low memory overhead, real-time upserts with primary key deduplication, and time series aggregations on both streaming and finite dimensional data.

History

Developed by Uber to meet their specific need "to make similar queries over relatively small, yet highly valuable, subsets of data (with maximum data freshness) at high QPS and low latency," with queries such as time series aggregations over geofences.

Checkpoints

Fuzzy

Snapshots are triggered by either a certain number of mutations or a certain time frame specific to each table.

Compression

Run-Length Encoding

Only compresses data with user defined sort orders that are low cardinality.

Data Model

Relational

Foreign Keys

Supported

AresDB supports foreign key joins.

Hardware Acceleration

GPU

Uses GPUs for query execution

Indexes

Hash Table

Used mainly for primary key deduplication.

Joins

Hash Join

Supports hash joins from fact tables (finite set data such as cities) to dimension tables (infinite streaming data such as rides). Also supports geospatial joins (i.e. geographically bounded area overlap) and normal foreign key joins. Note that AresDB uses late materialization for its joins, meaning the join may not be executed until a foreign key is accessed.

Logging

Logical Logging

Log files contain description of database upserts which must be replayed to rebuild the database after a crash.

Parallel Execution

Intra-Operator (Horizontal)

Executes queries with the one operation per kernel (OOPK) model.

Query Compilation

Not Supported

Query Execution

Vectorized Model

AresDB works with vector batches that are efficiently processed in parallel using the Thrust library.

Query Interface

Custom API

Uses a proprietary execution language called Ares Query Language (AQL) which is based in the JSON format, making it compatible with any language that can handle files and/or JSON.

Storage Architecture

Hybrid

Data within the archival delay of a table is kept uncompressed in live batches, while everything else is stored in compressed archival batches. If new data is ingested that is outside the archival array, it's added to an archival backfill queue which will be inserted into the archived batches asynchronously.

Storage Organization

Sorted Files

Archived data is sorted in a user specified column order, and files are organized by UTC day and Unix time cutoffs.

Stored Procedures

Not Supported

System Architecture

Shared-Disk

The CPU is only used to load information from storage into CPU memory and to distribute this data to GPU memory. The database system delegates each operator in a query to some GPU, so it's able to handle multiple GPUs by delegating different operations to different GPUs, each of which have completely separate memory. There are plans to implement proper distributed designs, but currently we're limited to a single system with multiple GPUs.

Views

Materialized Views

AresDB uses late materialization for its joins, meaning that it may only physically execute the join once a foreign key is accessed.

People Also Viewed

AresDB Logo
Website

https://eng.uber.com/aresdb/

Source Code

https://github.com/uber/aresdb

Tech Docs

https://github.com/uber/aresdb/wiki

Developer

Uber

Country of Origin

US

Start Year

2018

Project Type

Open Source

Written in

C, C++, Go

Inspired By

Elasticsearch, Kinetica, Ocelot, OmniSci, Pinot

Operating Systems

Linux

Licenses

Apache v2

People Also Viewed