4.8 KiB
Advice for developers new to operations
This section introduces some operational concepts and tasks to developers who have not written cloud applications before.
Monitoring
Monitoring is essential for 'scalable' cloud applications. You must
know how many requests are coming in and the impact that these requests
have on various services. You must have enough information to determine
whether to start another worker or API service as you did in /scaling_out.
explain how to achieve this kind of monitoring. Ceilometer? (STOP LAUGHING.)
In addition to this kind of monitoring, you should consider availability monitoring. Although your application might not care about a failed worker, it should care about a failed database server.
Use the Health Endpoint Monitoring Pattern <https://msdn.microsoft.com/en-us/library/dn589789.aspx> to implement functional checks within your application that external tools can access through exposed endpoints at regular intervals.
Backups
Just as you back up information on a non-cloud server, you must back up non-reproducible information, such as information on a database server, file server, or in application log files. Just because something is 'in the cloud' does not mean that the underlying hardware or systems cannot fail.
OpenStack provides a couple of tools that make it easy to back up data. If your provider runs OpenStack Object Storage, you can use its API calls and CLI tools to work with archive files.
You can also use the OpenStack API to create snapshots of running instances and persistent volumes. For more information, see your SDK documentation.
Link to appropriate documentation, or better yet, link and also include the commands here.
In addition to configuring backups, review your policies about what you back up and how long to retain each backed up item.
Phoenix servers
Phoenix Servers, named for the mythical bird that is consumed by fire and rises from the ashes to live again, make it easy to start over with new instances.
Application developers and operators who use phoenix servers have access to systems that are built from a known baseline, such as a specific operating system version, and to tooling that automatically builds, installs, and configures a system.
If you deploy your application on a regular basis, you can resolve outages and make security updates without manual intervention. If an outage occurs, you can provision more resources in another region. If you must patch security holes, you can provision additional compute nodes that are built with the updated software. Then, you can terminate vulnerable nodes and automatically fail-over traffic to the new instances.
Security
If one application instance is compromised, all instances with the same image and configuration will likely suffer the same vulnerability. The safest path is to use configuration management to rebuild all instances.
Configuration management
Configuration management tools, such as Ansible, Chef, and Puppet, enable you to describe exactly what to install and configure on an instance. Using these descriptions, these tools implement the changes that are required to get to the desired state.
These tools vastly reduce the effort it takes to work with large numbers of servers, and also improve the ability to recreate, update, move, and distribute applications.
Application deployment
How do you deploy your application? For example, do you pull the latest code from a source control repository? Do you make packaged releases that update infrequently? Do you perform haphazard tests in a development environment and deploy only after major changes?
One of the latest trends in scalable cloud application deployment is continuous integration and continuous deployment (CI/CD).
CI/CD means that you always test your application and make frequent deployments to production.
In this tutorial, we have downloaded the latest version of our application from source and installed it on a standard image. Our magic installation script also updates the standard image to have the latest dependencies that you need to run the application.
Another approach is to create a 'gold' image, which pre-installs your application and its dependencies. A 'gold' image enables faster boot times and more control over what is on the instance. However, if you use 'gold' images, you must have a process in place to ensure that these images do not fall behind on security updates.
Fail fast
Section needs to be written.