Quartz server can be configured clustered to have multiple servers run the same set of cron jobs. It’s good for load balancing and failover. Each quartz job can be set an option ‘request recovery’, which means whenever the job crashed (due to server down, application server down etc), the job can be recovered by other alive instances in the cluster.

More and more teams go with data-driven IT operations, for its rapid feedback cycle and effective iteration. Metrics, which can collect and display data in different forms, is becoming more and more important. If we want to provide Metrics as a Service, what do we need?

One of my colleagues in DevOps team creates a tool named EOM in my previous company. It’s basically an orchestration framework to pipeline the tasks needed in development environments provisioning. It handles failures, cocurrently run multiple tasks. However, the framework was only used inside the DevOps team, for this specific task. There should be broader scenarios for the framework to be used.

Rancher is an open source software platform that implements a purpose-built infrastructure for running Docker containers in production. As in production, there are requirements to create new requirements in infrastructure services such as networking, storage, load balancer, security, service discovery, and resource management.