Database Reference
In-Depth Information
Cloudstone benchmark
M / N
satisfies pre-defined read/write ratio
The controller
Monitor and
manage to scale
Replication within and across regions
Master
MySQL Proxy
M / N read
write split
Slave
us-west-1
Slave
us-west-2
Slave
us-east-1
Slave
us-east-2
Slave
eu-west-1
Slave
eu-west-2
Fig. 7.3
The implementation of the SLA management framework in the setup of experiments
The experiment setup is a multiple-layer implementation. The first layer rep-
resents the Cloudstone benchmark which generates an increasing workload of
database requests with a fixed read/write ratio. The benchmark is deployed in
a large instance to avoid any overload on the application tier. The second layer
hosts the MySQL Proxy and the SLA management framework. MySQL Proxy
with read and write split enabled resides in the middle between the benchmark
and the virtualized database replica servers, and acts as a load balancer to forward
read and write operations to the master and slaves correspondingly. The third layer
represents the database tier that consists of all the replica servers where the master
database receives the write operations from the load balancer after which it becomes
responsible for propagating the
writesets
to all the virtualized database slave servers.
The master database runs in a small instance so that an increasing replication delay
is expected to be observed along with an increasing workload. The master database
is closely located to the benchmark, the load balancer and the SLA management
framework. They are all deployed in the location of
us-west
. The slave servers are
responsible for serving the read operations and updating the writesets. They are
