We are developing an SaaS application, that will serve the mass market. We want a server administrator, AWS expert to setup an autogrowing server infrastructure in AWS. This means he will setup the Linux server, web, mail server on it and all needed software. Take care of all security. All services like Load Balancers, EC2, S3, DynamoDB, Route 5s, AutoScaling should be setup in order for the site to start automatically new EC2 servers when the traffic is high, and stop them when it gets low. The freelancer also should install and setup VoIP software like Asterix or other for 5 softphone accounts.
One development EC2 instance with all requirements for development server should be setup: Webserver, Database, PHP, Continuous Integration Server, SVN or Git Server. VPN should be created to be the only way to connect to the FTP of the production or development server. Other securities should also be applied.
The bid should also include 2 hours training and advises.
After the project a new maintenance agreement can be done on monthly bases.
Don't mind the budget specified, place your price whatever it is. The work has to be 100% perfect, so you place your bid only if you are good for the job.
I have been working as a linux system administrator over 9 years and have extensive experiences in nearly all fields of IT infrastructure setups and management.
please kindly take a minute to review my full message sent as private message.
thank you
I would prefer to suggest SVN as that is the only one that I have used in the past both as user and administrator. And Hudson or Jenkins for the continuous integration system. The best architecture would be to create two VPC in the AWS infrastructure both with dual subnet. Updates to the software using ftp would not be recommendable as deploying on all dynamic instances would be herculian, instead the continuous integration could be configured for a SVN build which could in turn update all running instances, whereas instances which startup from the predefined AMI could auto update themselves from the tagged SVN. The latest project which I did for a local body was hosted on AWS, and was having of the order of 3000 simultaneous users, with a ramp up pattern of 30 - 3000 - 100 during morning - noon - evening every day. And the system was achieved by utilizing a mix of spot and on demand instances using three different cloud watch alarms. Also please note that the scope does not cover any code modification to adopt system for the AWS, though ethically I may extend such suggestions.