5.3 Summarize orchestration processes and concepts Flashcards
What are agent monitoring tools?
Agent monitoring tools are orchestration utilities that require software (an agent) to be installed in the app container being monitored. These agents collect the data and transmit it to another location, such as a monitor server. The monitor server manages the information, provides analysis reporting, and also sends alerts for events, such as a container crashing
Describe agentless monitoring tools?
Agentless monitoring tools are also orchestration utilities. In this case, an agent is not installed in the app container being monitored. Instead, the tool uses preexisting and/or embedded software in the container or the container’s external environment to conduct its monitoring activity
Describe orchestration procedures
An app container is first designed with a chosen operating system, libraries, and supporting software installed as security and authentication policies. After the image design is complete, it is documented and stored, typically in a version control registry. The app container is deployed and replicated as needed. At this point, the app container moves into the monitoring phase, where it continues until the container images are no longer needed
Explain orchestrated app container attributes
App container images are static (immutable) and do not change after they have been designed and documented. For a development app container, only the app is modified, not the infrastructure or policies. A container image is registered with a version control system to assist in troubleshooting as well as rollbacks, if needed. When a new app container image is ready, the old production container image is stopped, not updated. The new app container is started in its place. App containers can be replicated. Thus, if additional app containers are needed, the new image is replicated to meet the needs.
Summarize container monitoring
Orchestration monitoring tools can have software installed on the app container (agent) or use already embedded or other preexisting container software (agentless). The monitoring utilities gather performance data and software metrics and watch the app containers’ health. If the container inventory should drop below the desired state, the orchestration tool can use automated configuration management and build automation to replicate the number of containers to bring the app container inventory back to the desired state
What is the purpose of orchestration?
The purpose of orchestration is to simplify complex setup tasks. Often, setting up software or a service requires multiple steps, and sometimes these steps need to be performed on multiple systems. With orchestration, these steps are planned out in advance and performed automatically, either by running a single command or pressing a button on a GUI or web-based interface.
Often a component of orchestration is monitoring. Monitoring is used to determine if an action needs to take place due to something that has happened on a system. There are two types of monitoring: agent and agentless
What is an agent?
An agent monitoring solution means that some software piece has been installed on the system or component that is being monitored. This agent actively probes the system, determining based on these probes if changes have been made that require an action
What is agentless?
An agentless monitoring solution means that no software has been installed on the system. An agentless system may either receive data from the system or conduct remote queries to determine if a change has taken place
What is a procedure?
A procedure for orchestration is the collection of steps that need to take place to complete the action. A procedure can be a fairly straightforward collection of steps, or it can be very complex and tied to the values of attributes
What are attributes?
An attribute is used to define parameters that are used to customize the automation procedure. These attributes contain data that can (and likely will) be different for each orchestration process. For example, suppose the procedure is designed to set up a system that requires access to the network. Attributes would need to be defined for things like the system’s IP address, subnet mask, gateway, and router
Describe Infrastructure Automation
Infrastructure automation is a component of orchestration. While many people use the terms orchestration and automation interchangeably, there is a difference between the two. You can certainly have orchestration without automation, but often a primary goal of orchestration is to automate the process.
Consider the following orchestration scenario: A user fills in fields of data in a web-based form. The data is used to populate attributes (see “Attributes” in this chapter). The user then clicks an “OK” button, and a procedure (see “Procedures” in this chapter) is performed to orchestrate a service. This is orchestration without automation.
Now consider this scenario: An agent is monitoring a system when a threshold is reached (say, maximum CPU usage in an hour). This triggers a procedure using attributes that are populated automatically from a database to spawn a second instance of the services. No human interaction took place; the orchestration process handed everything automatically. This is orchestration with automation.
Infrastructure as code
Infrastructure as Code (also referred to as IaC) is the heart of the orchestration process. Instead of manually configuring or provisioning systems, IaC makes use of software tools (orchestration products) to perform these tasks
Inventory
When something is orchestrated, the orchestration software needs to keep track of it. Often the “something” is referred to as an instance, and the information regarding this instance is stored in a database called an inventory
Automated Configuration Management
Configuration management has long been a standard process that system administrators have manually handled. Because it was not automated, this process was often time-consuming and resulted in misconfigured systems.
The purpose of automated configuration management is to ensure that software, services, and systems are maintained in a consistent, predictable state. By using automated configuration management tools, a large amount of the workload is handled by automated processes. This also results in fewer errors when compared to manual configuration management operations.
Common configuration management tools include the following:
Chef
Puppet
Ansible
Build Automation
A build (also called a software build) is when new source code is built into executable code, typically through a process called compiling. This can be a tedious process and subject to errors when performed manually, as there can be many complex steps in building large programs.
Build automation makes use of a utility that follows a set of rules to automate the software build process. When configured correctly, build automation can speed up the build process as well as reduce errors that occur when this process is performed manually.
Here are some common build automation tools:
Make
Ant
Gradle
Rake
Cake
