Deploying on Amazon Web Services

Benefits

• AWS provides automatic load balancing and failover between the tracker instances
• Directly write to S3 without further costs for further analysis (including by using Elastic MapReduce)

Quick steps

• Create a S3 bucket
• Create an IAM role with restricted access, only on this S3 bucket
• Get the Access and Secret keys for this IAM role
• Deploy at least two EC2 instances. Prefer high CPU instances, like c1.medium
• Install WT1 on the two instances
  • Select S3 storage
  • Enter the access and secret keys retrieved earlier
  • Don't forget, in S3 storage, to set an instance id for each EC2 instance. See Storage engines for more details
• Install nginx on the two instances
• Setup nginx to forward traffic to WT1. See Deployment with nginx
  • Get nginx to forward on instance_name:80/ --> wt1/public
• Create an Elastic Load Balancer
  • Use /isAlive as Health Check
• Add a CNAME to your Elastic Load Balancer host
• Set this CNAME as the tracking URL in the track.js