A companion guide to the ForgeRock® Cloud Deployment Model Cookbook for GKE. Provides guidance for customizing the Cloud Deployment Model (CDM) in a production deployment on Google Kubernetes Engine (GKE).
The Site Reliability Guide for GKE provides instructions to help you keep ForgeRock Identity Platform™ running smoothly on GKE. The guide contains information about monitoring, securing, backing up, and restoring the CDM.
Before You Begin
Before deploying the ForgeRock Identity Platform in a DevOps environment, read the important information in Start Here.
About ForgeRock Identity Platform Software
ForgeRock Identity Platform™ serves as the basis for our simple and comprehensive Identity and Access Management solution. We help our customers deepen their relationships with their customers, and improve the productivity and connectivity of their employees and partners. For more information about ForgeRock and about the platform, see https://www.forgerock.com.
Chapter 1. Monitoring Your Deployment
This chapter describes the CDM's monitoring architecture. It also covers common customizations you might perform to change the way monitoring, reporting, and sending alerts works in your environment.
1.1. About CDM Monitoring
The CDM uses Prometheus to monitor ForgeRock Identity Platform components and Kubernetes objects, Prometheus Alertmanager to send alert notifications, and Grafana to analyze metrics using dashboards.
Prometheus and Grafana are deployed when you install Helm charts from the
project into the
monitoring namespace of a CDM
cluster. These Helm charts deploy Kubernetes pods that run the Prometheus
and Grafana services.
The following Prometheus and Grafana pods from the
prometheus-operator project run in the
Handles Prometheus alerts by grouping them together, filtering them, and then routing them to a receiver, such as a Slack channel.
Generates Prometheus metrics for cluster node resources, such as CPU, memory, and disk usage. One pod is deployed for each CDM node.
Generates Prometheus metrics for Kubernetes API objects, such as deployments and nodes.
Provides the Grafana service.
Enables Prometheus configuration through custom Kubernetes objects.
Provides the Prometheus service.
kube-prometheus project README file
for more information about the pods in the preceding table.
In addition to the pods from the
import-dashboards-... pod from the
project runs after Grafana starts up. This pod imports Grafana
dashboards from the ForgeRock Identity Platform and terminates after importing has
To enable monitoring in the CDM, see
To access CDM monitoring dashboards, see:
The CDM uses Prometheus and Grafana for monitoring, reporting, and sending alerts. If you prefer to use different tools, deploy infrastructure in Kubernetes to support those tools.
Prometheus and Grafana are evolving technologies. Descriptions of these technologies were accurate at the time of this writing, but might differ when you deploy them.
1.2. Importing Custom Grafana Dashboards
The CDM includes a set of Grafana dashboards. You can customize, export and import Grafana dashboards using the Grafana UI or HTTP API.
For information about importing custom Grafana dashboards, see the
Import Custom Grafana Dashboards section in the
Prometheus and Grafana Deployment README file
1.3. Modifying the Prometheus Operator Configuration
The CDM's monitoring framework is based on the Prometheus Operator for Kubernetes project. The Prometheus Operator project provides monitoring definitions for Kubernetes services and deployment, and management of Prometheus instances.
When deployed, the Prometheus Operator watches for ServiceMonitor CRDs—Kubernetes Custom Resource Definitions. CRDs are Kubernetes class types that you can manage with the kubectl command. The ServiceMonitor CRDs define targets to be scraped.
In the CDM, the Prometheus Operator configuration is defined in the
files in the
repository. To customize the CDM monitoring framework, change values
in these files, following the examples documented in README files in the
Prometheus Operator project on GitHub.
1.4. Configuring Additional Alerts
To configure additional alerts, see the
Configure Alerting Rules
section in the Prometheus and Grafana Deployment README file
Chapter 2. Backing Up and Restoring Directory Data
This chapter describes the backup and restore functionality in CDM. It explains how to customize backup and restore operations in your deployment.
This chapter includes the following sections:
2.1. About Backup and Restore
The following diagram shows the backup topology of DS used in CDM.
Following are some important backup and restore considerations, as shown in the diagram above.
Three DS instances are deployed using Kubernetes stateful sets. Each stateful set is associated with a dedicated, persistent disk volume with a persistent volume claim (PVC). Each DS instance is a combined directory and replication server.
A shared-access file storage service, such as Google Cloud Filestore or Amazon EFS, is used to hold backups. The backup volume is mounted as a shared NFS volume using a PVC, and is used for storing backups of directory data. The mount point for the backup volume is the
/opt/opendj/bakdirectory in each DS instance.
dsadminpod stores and retrieves the DS backups.
The scripts necessary for performing backup and restore operations are available in the
scriptsfolder of each directory server pod.
You can use one of the DS pods to schedule backups using the kubectl exec command. See "Performing User-Initiated Backups" for more information on how to enable and schedule backups.
The backup schedule is configured based on the load and on a purge delay of 8 hours for DS instances. The default backup schedule is configured for the following:
An incremental backup is performed every hour.
A full backup is performed once a day.
You can customize the schedule of backups based on your volume of activity and recovery objectives. To customize the backup schedule see "Customizing the Backup Schedule".
When you perform a restore operation on a CDM instance, the latest backup available in the
/opt/opendj/bakis used for restoring the CDM instance.
Optionally, you can archive backup to cloud storage.
2.2. Enabling CDM-Scheduled Backups
After you deploy the CDM, backups will not be made until you schedule them.
To schedule backups, run the schedule-backup.sh script in each DS pod to be backed up. The default backup schedule creates a full backup every day at 12:00 midnight, and an incremental backup every hour.
To start making backups using the default schedule, perform these steps:
$ kubectl exec userstore-0 -it schedule-backup.sh
$ kubectl exec ctsstore-0 -it schedule-backup.sh
$ kubectl exec configstore-0 -it schedule-backup.sh
2.3. Customizing the Backup Schedule
You can alter the backup schedule using the following procedure:
Log in to the DS instance using kubectl exec command. For example:
$ kubectl exec userstore-0 -it bash
schedule-backup.shfile in the
scriptsdirectory to match your required schedule, and change the following two lines in the
schedule-backup.shfile to suit your schedule:
FULL_CRON="0 0 * * *"
INCREMENTAL_CRON="0 * * * *"
For example the following two lines schedule a full backup every day at 3:00 AM, and an incremental backup every 30 minutes:
FULL_CRON="0 3 * * *"
INCREMENTAL_CRON="*/30 * * * *"
Log out of the DS instance and run the kubectl exec DS -it schedule-backup.sh command. For example:
$ kubectl exec userstore-0 -it schedule-backup.sh
The schedule-backup.sh command stops any previously scheduled backup jobs before initiating the new schedule.
2.4. Performing User-Initiated Backups
In addition to scheduled backups, you might want to make occasional backups—for example, before applying security patches. Backups made outside of the scope of scheduled backups are called user-initiated backups.
The scripts necessary to perform user-initiated backups are available in the
scripts folder of the Directory Services pods.
For more information on backing up DS data, see the
Backing Up Directory Data
section of the ForgeRock Directory Server Administration Guide.
To make a full backup of user store directory data on demand, use the following kubectl exec command:
$ kubectl exec userstore-0 -it -- scripts/backup.sh --start 0 Starting backup to /opt/opendj/bak/default/userstore-prod Backup task 20181127191506671 scheduled to start immediately . . . [27/Nov/2018:19:15:06 +0000] severity="NOTICE" msgCount=56 msgID=org.opends.messages.tool-283 message="The backup process completed successfully" [27/Nov/2018:19:15:06 +0000] severity="NOTICE" msgCount=57 msgID=org.opends.messages.backend-414 message="Backup task 20181127191506671 finished execution in the state Completed successfully" Backup task 20181127191506671 has been successfully completed
Verify that the backups you made are available on your local backup volume by using the list-backup.sh script:
$ kubectl exec userstore-0 -it -- scripts/list-backup.sh Listing backups in /opt/opendj/bak/default/userstore-prod adminRoot backups Backup ID: 20181127191506Z Backup Date: 27/Nov/2018:19:15:06 +0000 Is Incremental: false Is Compressed: true Is Encrypted: false Has Unsigned Hash: false Has Signed Hash: false Dependent Upon: none ads-truststore backups Backup ID: 20181127191506Z Backup Date: 27/Nov/2018:19:15:06 +0000 Is Incremental: false Is Compressed: true Is Encrypted: false Has Unsigned Hash: false Has Signed Hash: false Dependent Upon: none amIdentityStore backups Backup ID: 20181127191506Z Backup Date: 27/Nov/2018:19:15:06 +0000 Is Incremental: false Is Compressed: true Is Encrypted: false Has Unsigned Hash: false Has Signed Hash: false Dependent Upon: none ... tasks backups Backup ID: 20181127191506Z Backup Date: 27/Nov/2018:19:15:06 +0000 Is Incremental: false Is Compressed: true Is Encrypted: false Has Unsigned Hash: false Has Signed Hash: false Dependent Upon: none
2.5. Exporting User Data
Exporting directory data to LDIF is very useful to:
Back up directory data
Copy or move directory data from one directory to another
Initialize a new directory server
Use the following steps to export user data with the export-ldif.sh command. For more information on exporting directory data, see Importing and Exporting Directory Data in the ForgeRock Directory Services Administration Guide.
Run the export-ldif command in the Directory Services pod. For example:
$ kubectl exec userstore-0 -it -- scripts/export-ldif.sh Exporting LDIF Exporting ldif of adminRoot to /opt/opendj/bak/default/userstore-prod/adminRoot-112719262018.04.ldif Export task 20181127192605510 scheduled to start immediately [27/Nov/2018:19:26:05 +0000] severity="NOTICE" msgCount=0 msgID=org.opends.messages.backend-413 message="Export task 20181127192605510 started execution" [27/Nov/2018:19:26:05 +0000] severity="INFORMATION" msgCount=1 msgID=org.opends.messages.tool-1662 message="Exporting to /opt/opendj/bak/default/userstore-prod/adminRoot-112719262018.04.ldif" [27/Nov/2018:19:26:05 +0000] severity="NOTICE" msgCount=2 msgID=org.opends.messages.backend-414 message="Export task 20181127192605510 finished execution in the state Completed successfully" Export task 20181127192605510 has been successfully completed
Run the ls command in the DS pod and verify that the LDIF files have been created on the
$ kubectl exec userstore-0 -it -- ls -l ./bak/default/userstore-prod/ total 48 drwxr-xr-x 2 forgerock forgerock 4096 Nov 27 19:15 adminRoot -rw------- 1 forgerock forgerock 2802 Nov 27 19:26 adminRoot-112719262018.04.ldif drwxr-xr-x 2 forgerock forgerock 4096 Nov 27 19:15 ads-truststore ...
2.6. Archiving the CDM Backup
To avoid losing backup files when a disaster occurs, archive the backup files
to Google Cloud Storage (GCS). First, configure a GCS bucket in your
deployment. Then, enable archiving backup files to GCS. You can configure the
dsadmin pod to archive your local backups to your GCS bucket.
The following diagram shows how backup files are archived to GCS:
The following procedure enables archiving all the backup files that have
been made for the
Set the following values in the YAML file that you use to install the
gcs.bucketto the URL for your storage bucket
Here is a sample
dsadmin.yamlfile snippet in which archive is enabled, and the backup is archived to a GCS bucket at the location gs://my-br-bucket:
gcs: enabled: true bucket: gs://my-br-bucket ... createPVC: false ...
$ helm delete --purge prod-dsadmin These resources were kept due to the resource policy: [PersistentVolumeClaim] ds-backup [PersistentVolume] ds-backup release "prod-dsadmin" deleted
$ helm install --name prod-dsadmin --namespace prod \ --values common.yaml --values dsadmin.yaml /path/to/forgeops/helm/dsadmin NAME: prod-dsadmin LAST DEPLOYED: Fri Nov 30 11:13:10 2018 NAMESPACE: prod STATUS: DEPLOYED RESOURCES: ==> v1beta1/CronJob NAME AGE gcs-sync 1s ==> v1/Pod(related) NAME READY STATUS RESTARTS AGE dsadmin-784fd647b4-lcgjg 0/1 Init:0/1 0 1s ==> v1/Deployment NAME AGE dsadmin 1s
You can view the list of backups archived to your GCS bucket by using the Google Cloud Platform UI.
For more information about the backup archive process, see the
README.md file for the
dsadmin Helm chart.
2.7. Using CDM Restore
You can enable the restore capability in the CDM, and redeploy DS pods, to restore data from the latest available backup. This is useful when a system disaster occurs, or when directory services are lost. This is also useful when you want to port test environment data to a production deployment. Note that when you redeploy DS data, the entire DS environment is restored. To restore specific portions of the directory data without having to redeploy the DS environment, see "User-Initiated Restore".
Use these steps to redeploy the user store from a backup:
If either the
userstore-1pod is running, list the backups available. For example, if the
userstore-0pod is running, execute the following command:
$ kubectl exec userstore-0 -it -- scripts/list-backup.sh Listing backups in /opt/opendj/bak/default/userstore-prod adminRoot backups Backup ID: 20181129202519Z Backup Date: 29/Nov/2018:20:25:27 +0000 Is Incremental: false Is Compressed: true Is Encrypted: false Has Unsigned Hash: false Has Signed Hash: false Dependent Upon: none ads-truststore backups Backup ID: 20181129202519Z Backup Date: 29/Nov/2018:20:25:27 +0000 Is Incremental: false Is Compressed: true Is Encrypted: false Has Unsigned Hash: false Has Signed Hash: false Dependent Upon: none ... tasks backups Backup ID: 20181129202519Z Backup Date: 29/Nov/2018:20:25:27 +0000 Is Incremental: false Is Compressed: true Is Encrypted: false Has Unsigned Hash: false Has Signed Hash: false Dependent Upon: none
$ helm delete --purge prod-userstore release "prod-userstore" deleted
List the PVCs in your namespace using the kubectl get pvc command:
$ kubectl get pvc NAME STATUS VOLUME CAPACITY ACCESS MODES STORAGECLASS AGE db-configstore-0 Bound pvc-2a1caac1-f3c8-11e8-8de8-42010a8e005c 10Gi RWO standard 9h db-configstore-1 Bound pvc-5f517dab-f3c8-11e8-8de8-42010a8e005c 10Gi RWO standard 9h db-ctsstore-0 Bound pvc-65f4af80-f3c8-11e8-8de8-42010a8e005c 100Gi RWO fast 9h db-ctsstore-1 Bound pvc-82c993f2-f3c8-11e8-8de8-42010a8e005c 100Gi RWO fast 9h db-userstore-0 Bound pvc-478ef730-f3c8-11e8-8de8-42010a8e005c 100Gi RWO fast 9h db-userstore-1 Bound pvc-6dc544a6-f3c8-11e8-8de8-42010a8e005c 100Gi RWO fast 9h ds-backup Bound ds-backup 1Ti RWX nfs 9h
Delete the PVCs for userstore using the kubectl delete pvc command:
$ kubectl delete pvc db-userstore-1 persistentvolumeclaim "db-userstore-1" deleted $ kubectl delete pvc db-userstore-0 persistentvolumeclaim "db-userstore-0" deleted
Do not delete the
ds-backupPVC. If you delete the backup PVC, you cannot restore data.
userstore.yamlfile, and set the
... # Restore parameters. restore: enabled: true ...
userstoredirectory. During deployment, the directory data is restored from the backup:
$ helm install --name prod-userstore --namespace prod \ --values common.yaml --values userstore.yaml /path/to/forgeops/helm/ds NAME: prod-userstore LAST DEPLOYED: Thu Nov 29 13:04:00 2018 NAMESPACE: prod STATUS: DEPLOYED RESOURCES: ==> v1/Secret NAME AGE userstore 0s ==> v1/ConfigMap userstore 0s ==> v1/Service userstore 0s ==> v1beta1/StatefulSet userstore 0s ==> v1/Pod(related) NAME READY STATUS RESTARTS AGE userstore-0 0/1 Init:0/1 0 0s
Verify that the DS pods are all running:
$ kubectl get pods NAME READY STATUS RESTARTS AGE ... userstore-0 1/1 Running 0 11m userstore-1 0/1 Running 0 4m
2.8. User-Initiated Restore
To manually restore DS data, use a backup from the backup volume.
First, run the list-backup.sh to list the backups available
on your local backup volume. Then, run either the restore
command in the
bin directory, or the restore.sh
script in the
scripts directory to restore directory data.
For more information about the restore command and the
restore.sh script, see the
ForgeRock Directory Server
Consider following these best practices:
Use a backup that is newer than the last replication purge.
When you restore a DS replica using backups that are older than the purge delay, that replica will no longer be able to participate in replication. Reinitialize the replica to restore the replication topology.
If the available backups are older than the purge delay, then initialize the DS replica from an up-to-date master instance. For more information on how to initialize a replica from a master instance, see Initializing Replicas section of ForgeRock Directory Server Administration Guide.
Chapter 3. Securing Your Deployment
This chapter describes options for securing your ForgeRock Identity Platform deployment.
3.1. Securing Communication With ForgeRock Identity Platform Servers
The ForgeRock DevOps Examples and CDM enable secure communication with ForgeRock Identity Platform services using an SSL-enabled ingress controller. Incoming requests and outgoing responses are encrypted. SSL is terminated at the ingress controller.
You can configure communication with ForgeRock Identity Platform services using one of the following options:
Over HTTPS using a self-signed certificate. Communication is encrypted, but users will receive warnings about insecure communication from some browsers. For configuration steps, see "Configuring and Installing the frconfig Helm Chart" in the DevOps Developer's Guide.
Over HTTPS using a certificate with a trust chain that starts at a trusted root certificate. Communication is encrypted, and users will not receive warnings from their browsers. For configuration steps, see "Configuring and Installing the frconfig Helm Chart" in the DevOps Developer's Guide.
Over HTTPS using a dynamically obtained certificate from Let's Encrypt. Communication is encrypted and users will not receive warnings from their browsers. A
cert-managerpod installed in your Kubernetes cluster calls Let's Encrypt to obtain a certificate, and then automatically installs a Kubernetes secret. This option is covered in the next section.
You install a Helm chart from the
to provision certificates. By default, the pod issues a self-signed
certificate. You can also configure the pod to issue a certificate with a
trust chain that begins at a trusted root certificate, or to dynamically
obtain a certificate from Let's Encrypt.
3.1.1. Automating Certificate Management Using Let's Encrypt
In the CDM, certificate management is provided by the
cert-manager add-on. The certificate manager
deployed in CDM specifies Let's Encrypt as the certificate issuer
and Google Cloud DNS as the DNS01 challenge provider. This configuration
secures CDM communication with a dynamically obtained certificate
from Let's Encrypt.
For the steps the Cloud Deployment Team used to deploy the
cert-manager add-on in the CDM, see
"Deploying the Certificate Manager" in the Cloud Deployment Model Cookbook for GKE.
Appendix A. Getting Support
This appendix contains information about support options for the ForgeRock DevOps Examples and the ForgeRock Identity Platform.
A.1. Statement of Support
The ForgeRock DevOps Examples, the CDM, and the accompanying Git repository demonstrate deployment in a containerized environment using DevOps techniques. You are responsible for adapting the examples to suit your production requirements. These resources are provided for demonstration purposes only. Commercial support for the DevOps Examples and the CDM is not available from ForgeRock.
ForgeRock does not support production deployments that use the evaluation-only Docker images described in "Using the Evaluation-Only Docker Images" in the DevOps Developer's Guide. When deploying the ForgeRock Identity Platform by using Docker images, you must build and use your own images for production deployments. For information about how to build Docker images for the ForgeRock Identity Platform, see "Using the Evaluation-Only Docker Images" in the DevOps Developer's Guide.
ForgeRock provides commercial support for the ForgeRock Identity Platform only. For supported components, containers, and Java versions, see the following:
Before You Install in the ForgeRock Access Management Release Notes.
Before You Install in the ForgeRock Identity Management Release Notes.
Before You Install in the ForgeRock Directory Services Release Notes.
Before You Install in the ForgeRock Identity Message Broker Release Notes.
Before You Install in the ForgeRock Identity Gateway Release Notes.
ForgeRock does not provide support for software that is not part of the ForgeRock Identity Platform, such as Docker, Kubernetes, Java, Apache Tomcat, NGINX, Apache HTTP Server, and so forth.
A.2. Accessing Documentation Online
ForgeRock publishes comprehensive documentation online:
The ForgeRock Knowledge Base offers a large and increasing number of up-to-date, practical articles that help you deploy and manage ForgeRock software.
While many articles are visible to community members, ForgeRock customers have access to much more, including advanced information for customers using ForgeRock software in a mission-critical capacity.
ForgeRock product documentation, such as this document, aims to be technically accurate and complete with respect to the software documented. It is visible to everyone and covers all product features and examples of how to use them.
A.3. How to Report Problems or Provide Feedback
If you have questions regarding the DevOps Examples or the CDM that are not answered by the documentation, you can ask questions on the DevOps forum at https://forum.forgerock.com/forum/devops.
When requesting help with a problem, include the following information:
Description of the problem, including when the problem occurs and its impact on your operation.
Description of the environment, including the following information:
Environment type (Minikube, GKE, or EKS).
Software versions of supporting components:
Oracle VirtualBox (Minikube environments only).
Docker client (all environments).
Minikube (all environments).
kubectl command (all environments).
Kubernetes Helm (all environments).
Google Cloud SDK (GKE environments only).
Amazon AWS Command Line Interface (EKS environments only).
Any patches or other software that might be affecting the problem.
Steps to reproduce the problem.
HTML output from the debug-logs.sh script. For more information, see "Running the debug-logs.sh Script" in the DevOps Developer's Guide.
A.4. Getting Support and Contacting ForgeRock
ForgeRock provides support services, professional services, training through ForgeRock University, and partner services to assist you in setting up and maintaining your deployments. For a general overview of these services, see https://www.forgerock.com.
- affinity (AM)
AM affinity based load balancing ensures that the CTS token creation load is spread over multiple server instances (the token origin servers). Once a CTS token is created and assigned to a session, all subsequent token operations are sent to the same token origin server from any AM node. This ensures that the load of CTS token management is spread across directory servers.
- Amazon EKS
Amazon Elastic Container Service for Kubernetes (Amazon EKS) is a managed service that makes it easy for you to run Kubernetes on Amazon Web Services without needing to set up or maintain your own Kubernetes control plane.
- ARN (AWS)
An Amazon Resource Name (ARN) uniquely identifies an Amazon Web Service (AWS) resource. AWS requires an ARN when you need to specify a resource unambiguously across all of AWS, such as in IAM policies and API calls.
- AWS IAM Authenticator for Kubernetes
The AWS IAM Authenticator for Kubernetes is an authentication tool that enables you to use Amazon Web Services (AWS) credentials for authenticating to a Kubernetes cluster.
cloud-controller-managerdaemon runs controllers that interact with the underlying cloud providers.
cloud-controller-manageris an alpha feature introduced in Kubernetes release 1.6. The
cloud-controller-managerdaemon runs cloud-provider-specific controller loops only.
- Cloud Deployment Model (CDM)
The Cloud Deployment Model (CDM) is a common use ForgeRock Identity Platform architecture, designed to be easy to deploy and easy to replicate. The ForgeRock Cloud Deployment Team has developed Helm charts, Docker images, and other artifacts expressly to build the CDM.
- CloudFormation (AWS)
CloudFormation is a service that helps you model and set up your Amazon Web Services (AWS) resources. You create a template that describes all the AWS resources that you want. AWS CloudFormation takes care of provisioning and configuring those resources for you.
- CloudFormation template (AWS)
An AWS CloudFormation template describes the resources that you want to provision in your AWS stack. AWS CloudFormation templates are text files formatted in JSON or YAML.
A container cluster is the foundation of Kubernetes Engine. A cluster consists of at least one cluster master and multiple worker machines called nodes. The Kubernetes objects that represent your containerized applications all run on top of a cluster.
- cluster master
A cluster master schedules, runs, scales and upgrades the workloads on all nodes of the cluster. The cluster master also manages network and storage resources for workloads.
A configuration map, called
ConfigMapin Kubernetes manifests, binds the configuration files, command-line arguments, environment variables, port numbers, and other configuration artifacts to the assigned containers and system components at runtime. The configuration maps are useful for storing and sharing non-sensitive, unencrypted configuration information.
A container is an allocation of resources such as CPU, network I/O, bandwidth, block I/O, and memory that can be “contained” together and made available to specific processes without interference from the rest of the system.
A set of daemons, called
DaemonSetin Kubernetes manifests, manages a group of replicated pods. Usually, the daemon set follows an one-pod-per-node model. As you add nodes to a node pool, the daemon set automatically distributes the pod workload to the new nodes as needed.
A Kubernetes deployment represents a set of multiple, identical pods. A Kubernetes deployment runs multiple replicas of your application and automatically replaces any instances that fail or become unresponsive.
- deployment controller
A deployment controller provides declarative updates for pods and replica sets. You describe a desired state in a deployment object, and the deployment controller changes the actual state to the desired state at a controlled rate. You can define deployments to create new replica sets, or to remove existing deployments and adopt all their resources with new deployments.
- Docker Cloud
Docker Cloud provides a hosted registry service with build and testing facilities for Dockerized application images; tools to help you set up and manage host infrastructure; and application lifecycle features to automate deploying (and redeploying) services created from images.
- Docker container
A Docker container is a runtime instance of a Docker image. A Docker container is isolated from other containers and its host machine. You can control how isolated your container’s network, storage, or other underlying subsystems are from other containers or from the host machine.
- Docker daemon
The Docker daemon (
dockerd) listens for Docker API requests and manages Docker objects such as images, containers, networks, and volumes. A Docker daemon can also communicate with other Docker daemons to manage Docker services.
- Docker Engine
The Docker Engine is a client-server application with these components:
A server, which is a type of long-running program called a daemon process (the
A REST API, which specifies interfaces that programs can use to talk to the daemon and tell it what to do
A command-line interface (CLI) client (the
A Dockerfile is a text file that contains the instructions for building a Docker image. Docker uses the Dockerfile to automate the process of building a Docker image.
- Docker Hub
Docker Hub provides a place for you and your team to build and ship Docker images. You can create public repositories that can be accessed by any other Docker Hub user, or you can create private repositories you can control access to.
An image is an application you would like to run. A container is a running instance of an image.
- Docker image
A Docker image is a read-only template with instructions for creating a Docker container. Often, an image is based on another image, with some additional customization.
A Docker image includes the application code, a runtime engine, libraries, environment variables, and configuration files that are required to run the application.
An image is an application you would like to run. A container is a running instance of an image.
- Docker namespace
Docker namespaces provide a layer of isolation. When you run a container, Docker creates a set of namespaces for that container. Each aspect of a container runs in a separate namespace and its access is limited to that namespace.
PIDnamespace is the mechanism for remapping process IDs inside the container. Other namespaces such as net, mnt, ipc, and uts provide the isolated environments we know as containers. The user namespace is the mechanism for remapping user IDs inside a container.
- Docker registry
A Docker registry stores Docker images. Docker Hub and Docker Cloud are public registries that anyone can use, and Docker is configured to look for images on Docker Hub by default. You can also run your own private registry.
- Docker repository
A Docker repository is a public, certified repository from vendors and contributors to Docker. It contains Docker images that you can use as the foundation to build your applications and services.
- Docker service
In a distributed application, different pieces of the application are called “services.” Docker services are really just “containers in production.” A Docker service runs only one image, but it codifies the way that image runs including which ports to use, the number replicas the container should run, and so on. By default, the services are load-balanced across all worker nodes.
- dynamic volume provisioning
The process of creating storage volumes on demand is called dynamic volume provisioning. Dynamic volume provisioning allows storage volumes to be created on-demand. It automatically provisions storage when it is requested by users.
An egress controls access to destinations outside the network from within a Kubernetes network. For an external destination to be accessed from a Kubernetes environment, the destination should be listed as an allowed destination in the whitelist configuration.
- firewall rule
A firewall rule lets you allow or deny traffic to and from your virtual machine instances based on a configuration you specify. Each Kubernetes network has a set of firewall rules controlling access to and from instances in its subnets. Each firewall rule is defined to apply to either incoming glossary-ingress(ingress) or outgoing (egress) traffic, not both.
- garbage collection
Garbage collection is the process of deleting unused objects. Kubelets perform garbage collection for containers every minute and garbage collection for images every five minutes. You can adjust the high and low threshold flags and garbage collection policy to tune image garbage collection.
- Google Kubernetes Engine (GKE)
The Google Kubernetes Engine (GKE) is an environment for deploying, managing, and scaling your containerized applications using Google infrastructure. The GKE environment consists of multiple machine instances grouped together to form a container cluster.
An ingress is a collection of rules that allow inbound connections to reach the cluster services.
- instance group
An instance group is a collection of instances of virtual machines. The instance groups enable you to easily monitor and control the group of virtual machines together.
- instance template
An instance template is a global API resource that you can use to create VM instances and managed instance groups. Instance templates define the machine type, image, zone, labels, and other instance properties. They are very helpful in replicating the environments.
The kubectl command-line tool supports several different ways to create and manage Kubernetes objects.
The Kubernetes controller manager is a process that embeds core controllers that are shipped with Kubernetes. Logically each controller is a separate process, but to reduce complexity, they are all compiled into a single binary and run in a single process.
A kubelet is an agent that runs on each node in the cluster. It ensures that containers are running in a pod.
kube-schedulercomponent is on the master node and watches for newly created pods that do not have a node assigned to them, and selects a node for them to run on.
Kubernetes is an open source platform designed to automate deploying, scaling, and operating application containers.
Source: Kubernetes Concepts
- Kubernetes DNS
A Kubernetes DNS pod is a pod used by the kubelets and the individual containers to resolve DNS names in the cluster.
- Kubernetes namespace
A Kubernetes namespace is a virtual cluster that provides a way to divide cluster resources between multiple users. Kubernetes starts with three initial namespaces:
default: The default namespace for user created objects which don't have a namespace
kube-system: The namespace for objects created by the Kubernetes system
kube-public: The automatically created namespace that is readable by all users
Kubernetes supports multiple virtual clusters backed by the same physical cluster.
- Let's Encrypt
Let's Encrypt is a free, automated, and open certificate authority.
Source: Let's Encrypt web site.
- network policy
A Kubernetes network policy specifies how groups of pods are allowed to communicate with each other and with other network endpoints.
- node (Kubernetes)
A Kubernetes node is a virtual or physical machine in the cluster. Each node is managed by the master components and includes the services needed to run the pods.
- node controller (Kubernetes)
A Kubernetes node controller is a Kubernetes master component that manages various aspects of the nodes such as: lifecycle operations on the nodes, operational status of the nodes, and maintaining an internal list of nodes.
- persistent volume
A persistent volume (PV) is a piece of storage in the cluster that has been provisioned by an administrator. It is a resource in the cluster just like a node is a cluster resource. PVs are volume plugins that have a lifecycle independent of any individual pod that uses the PV.
- persistent volume claim
A persistent volume claim (PVC) is a request for storage by a user. A PVC specifies size, and access modes such as:
Mounted once for read and write access
Mounted many times for read-only access
- pod anti-affinity (Kubernetes)
Kubernetes pod anti-affinity allows you to constrain which nodes can run your pod, based on labels on the pods that are already running on the node rather than based on labels on nodes. Pod anti-affinity enables you to control the spread of workload across nodes and also isolate failures to nodes.
- pod (Kubernetes)
A Kubernetes pod is the smallest, most basic deployable object in Kubernetes. A pod represents a single instance of a running process in a cluster. Containers within a pod share an IP address and port space.
- replication controller
A replication controller ensures that a specified number of Kubernetes pod replicas are running at any one time. The
replication controllerensures that a pod or a homogeneous set of pods is always up and available.
- secret (Kubernetes)
A Kubernetes secret is a secure object that stores sensitive data, such as passwords, OAuth 2.0 tokens, and SSH keys in your clusters.
- security group (AWS)
A security group acts as a virtual firewall that controls the traffic for one or more compute instances.
- service (Kubernetes)
A Kubernetes service is an abstraction which defines a logical set of pods and a policy by which to access them. This is sometimes called a microservice.
Sharding is a way of partitioning directory data so that the load can be shared by multiple directory servers. Each data partition, also known as a shard, exposes the same set of naming contexts, but only a subset of the data. For example, a distribution might have two shards. The first shard contains all users whose name begins with A-M, and the second contains all users whose name begins with N-Z. Both have the same naming context.
- stack (AWS)
A stack is a collection of AWS resources that you can manage as a single unit. You can create, update, or delete a collection of resources by using stacks. All the resources in a stack are defined by the template.
- stack set (AWS)
A stack set is a container for stacks. You can provision stacks across AWS accounts and regions by using a single AWS template. All the resources included in each stack of a stack set are defined by the same template.
- volume (Kubernetes)
A Kubernetes volume is a storage volume that has the same lifetime as the pod that encloses it. Consequently, a volume outlives any containers that run within the pod, and data is preserved across container restarts. When a pod ceases to exist, the Kubernetes volume also ceases to exist.
- VPC (AWS)
A virtual private cloud (VPC) is a virtual network dedicated to your AWS account. It is logically isolated from other virtual networks in the AWS Cloud.
- worker node (AWS)
An Amazon Elastic Container Service for Kubernetes (Amazon EKS) worker node is a standard compute instance provisioned in Amazon EKS.
- workload (Kubernetes)
A Kubernetes workload is the collection of applications and batch jobs packaged into a container. Before you deploy a workload on a cluster, you must first package the workload into a container.