azure-quickstart-templates/postgresql-on-ubuntu

Create multiple Ubuntu 14.04 VMs with PostgreSQL 9.3 master-slave replication and one jumpbox VM with a public IP

This template allows you to create one master PostgreSQL 9.3 server with streaming-replication to multiple (based on the 'numberOfSlaveInstances' parameter) slave servers in a private-only subnet. Each database server is configured with 2 data disks that are striped into RAID-0 configuration using mdadm. The template also creates one publicly accessible VM to serve as a jumpbox for ssh into the backend database servers.

The template creates the following deployment resources:

• Virtual Network with two subnets: Subnet-DMZ for the jumpbox VM and Subnet-DB for the PostgreSQL master and slave VMs
• Storage account to storage the VM VHDS
• Public IP address (named publicip) for accessing the jumpbox server via ssh
• Network Interface Card (NIC) for jumpbox VM (named nicapache)
• One Network Interface Card (NIC) for PostgreSQL master servers (named nicmaster)
• Multiple Network Interface Cards (NICs) for PostgreSQL slave servers (named nicslave0, nicslave1, etc.)
• Multiple remotely-hosted CustomScriptForLinux (install_postgresql.sh with passed in parameters) extensions to install and configure the PostgreSQL service on the master and each of the numberOfSlaveInstances VMs

NOTE: To access the PostgreSQL servers, you need to use the publicly accessble jumpbox VM and ssh from it into the backend servers.

Assuming your domainName parameter was "mypsqljumpbox", location was "West US", and subnetDbPrefix was 10.1.1.0/24:

• Master PostgreSQL server will be deployed at the first available IP address in the subnet: 10.1.1.4
• Slave PostgreSQL servers will be deployed in the other IP addresses: 10.1.1.5, 10.1.1.6, etc.
• From your computer, SSH into the jumpbox ssh mypsqljumpbox.westus.cloudapp.net
• From the jumpbox, SSH into the master PostgreSQL server ssh 10.1.1.4
• On the master (e.g. 10.1.1.4), use the following code to create table and some test data within your PostgreSQL master database.

sudo -u postgres psql
create table table1 (name varchar(100));
insert into table1 (name) values ('name1');
insert into table1 (name) values ('name2');
select * from table1;

• From the jumpbox, SSH into one of the slave PostgreSQL servers ssh 10.1.1.5 and use psql to check that the data propaged properly

sudo -u postgres psql
select * from table1;

Template expects the following parameters

Name	Description
region	Location where the resources will be deployed
storageAccountNamePrefix	Unique DNS name for the Storage Account where the Virtual Machines' disks will be placed
domainName	Domain name of the publicly accessible jumpbox VM {domainName}.{location}.cloudapp.com (e.g. mydomainname.westus.cloudapp.azure.com)
adminUsername	Username for the Virtual Machines
adminPassword	Password for the Virtual Machine
tshirtSize	Size of deployment to provision
replicatorPassword	Password to use for the pgsql replication user (replicator)
jumpbox	Enable jumpbox
virtualNetworkName	Virtual network name

The following table outlines the deployment topology characteristics for each supported t-shirt size:

T-Shirt Size	Database VM Size	CPU Cores	Memory	Data Disks	# of Secondaries	# of Storage Accounts
Small	Standard_A1	1	1.75 GB	2x1023 GB	1	1
Medium	Standard_A3	4	7 GB	8x1023 GB	1	2
Large	Standard_A4	8	14 GB	16x1023 GB	2	2
XLarge	Standard_A4	8	14 GB	16x1023 GB	3	4