Skip to content
svaroqui edited this page Feb 23, 2013 · 89 revisions

Welcome to the scrambledb wiki!

The primary Architecture of ScrambleDB is to provide a clustered database abstraction using open source products. The framework enable multi services managing, deploying, provisioning, hight availability, automatic and manual routing of SQL queries.

ARCHITECTURE OVERVIEW

Scramble DB Architecture Overview

Version 0.1

  • Hight avalability
  • Default policy of read consistency queries to slaves
  • Auto scale up on Amazon VPC
  • Auto scale out on Amazon VPC
  • Online upgrade to new release or configuration change
  • Multi cloud managment and local network tenancy not hybrid

Client or API named service command :

  • Services install

./clmgr services install bench1

  • Services start

./clmgr services start bench1

  • Services stop

./clmgr services stop proxy1

  • Services status

./clmgr services status proxy1

  • Services switch database master

./clmgr services switch db2

  • Services switch vip

./clmgr services switch_vip lb2

  • Services bootstrap_binaries : copy the full stack to all services IPs

./clmgr services bootstrap_binaries

  • Services bootstrap configuration : copy the etc/cloud.cnf to all services ip

./clmgr services bootstrap_config

  • Services bootstrap ncc : copy the cluster controler to all services ip

./clmgr services bootstrap_ncc

  • Services change configuration online

./clmgr services rolling_restart all db
./clmgr services rolling_restart all proxy

Client or API multi services command

  • Filtering per sevice mode

./clmgr services status all
./clmgr services status all mariadb

  • Filtering per localhost

./clmgr services status local all

Client or API cloud commands

  • Instances status

./clmgr instances status all

  • Instances launch

./clmgr instances launch ip

  • Instances start

./clmgr instances start ip

  • Instances stop

./clmgr instances stop ip

  • Instances terminate

./clmgr instances terminate ip

Features for Version 0.2

  • Cluster of services shard to multiple VIP DB abstraction

  • Relaxing consistency for sql queries

  • Routing of read sql queries to pool of slaves with table horizontal sharding

  • Routing of read sql queries to pool of slavesslaves with table or db vertical sharding queries

  • Routing of read sql queries to pool of sphinx RT indexes

  • Routing of read sql queries to pool of key value store

  • Routing of read sql queries to analytics slaves

Clone this wiki locally