In this lab you will explore the logging aggregation capabilities of OpenShift.
An extremely important function of OpenShift is collecting and aggregating logs from the environments and the application pods it is running. OpenShift ships with an elastic log aggregation solution: EFK. (ElasticSearch, Fluentd and Kibana)
The cluster logging components are based upon Elasticsearch, Fluentd, and Kibana (EFK). The collector, Fluentd, is deployed to each node in the OpenShift cluster. It collects all node and container logs and writes them to Elasticsearch (ES). Kibana is the centralized, web UI where users and administrators can create rich visualizations and dashboards with the aggregated data. Administrators can see and search through all logs. Application owners and developers can allow access to logs that belong to their projects. The EFK stack runs on top of OpenShift.
This lab requires that you have completed the infra-nodes lab. The logging
stack will be installed on the infra
nodes that were created in that lab.
More information may be found on the official OpenShift documentation site
This exercise is done almost entirely using the OpenShift web console. All of the interactions with the web console are effectively creating or manipulating API objects in the background. It is possible to fully automate the process and/or do it using the CLI or other tools, but these methods are not covered in the exercise or documentation at this time.
OpenShift Container Platform cluster logging is designed to be used with the default configuration, which is tuned for small to medium sized OpenShift Container Platform clusters. The installation instructions that follow include a sample Cluster Logging Custom Resource (CR), which you can use to create a cluster logging instance and configure your cluster logging deployment.
If you want to use the default cluster logging install, you can use the sample CR directly.
If you want to customize your deployment, make changes to the sample CR as needed. The following describes the configurations you can make when installing your cluster logging instance or modify after installtion. See the Configuring sections for more information on working with each component, including modifications you can make outside of the Cluster Logging Custom Resource.
OpenShift Logging will be run from within its own namespace
openshift-logging
. This namespace does not exist by default, and needs to
be created before logging may be installed. The namespace is represented in
yaml format as:
apiVersion: v1
kind: Namespace
metadata:
name: openshift-logging
annotations:
openshift.io/node-selector: ""
labels:
openshift.io/cluster-logging: "true"
openshift.io/cluster-monitoring: "true"
To create the namespace, run the following command:
oc create -f support/openshift_logging_namespace.yaml
Verify that it has been created:
oc get ns openshift-logging
You should see the following output:
NAME STATUS AGE
openshift-logging Active 11s
In order to install and configure the EFK
stack into the cluster,
additional operators need to be installed. These can be installed from the
Operator Hub
from within the cluster via the GUI.
When using operators in OpenShift, it is important to understand the basics
of some of the underlying principles that make up the Operators.
CustomResourceDefinion (CRD)
and CustomResource (CR)
are two Kubernetes
objects that we will briefly describe.CRDs
are generic pre-defined
structures of data. The operator understands how to apply the data that is
defined by the CRD
. In terms of programming, CRDs
can be thought as being
similar to a class. CustomResource (CR)
is an actual implementations of the
CRD
, where the structured data has actual values. These values are what the
operator will use when configuring it’s service. Again, in programming terms,
CRs
would be similar to an instantiated object of the class.
The general pattern for using Operators is first, install the Operator, which
will create the necessary CRDs
. After the CRDs
have been created, we can
create the CR
which will tell the operator how to act, what to install,
and/or what to configure. For installing openshift-logging
, we will follow
this pattern.
To begin, click on the Console
tab or use the following link to log-in
to the OpenShift Cluster’s GUI.
{{ MASTER_URL }}
Then follow the following steps:
-
Install the Elasticsearch Operator:
-
In the OpenShift console, click
Operators
→OperatorHub
. -
Type
Elasticsearch Operator
in the search field and click theOpenShift Elasticsearch Operator
card from the list of available Operators, and then clickInstall
. -
On the
Create Operator Subscription
page, select Update Channelstable
, leave all other defaults and then clickInstall
.This makes the Operator available to all users and projects that use this OpenShift Container Platform cluster.
-
-
Install the Cluster Logging Operator:
The
Cluster Logging
operator needs to be installed in theopenshift-logging
namespace. Please ensure that theopenshift-logging
namespace was created from the previous steps-
In the OpenShift console, click
Operators
→OperatorHub
. -
Type
OpenShift Logging
in the search box and click theRed Hat OpenShift Logging
card from the list of available Operators, and clickInstall
. -
On the
Install Operator
page, select Update Channelstable
. Under ensureInstallation Mode
ensure thatA specific namespace on the cluster
is selected, and chooseOperator recommended Namespace: openshift-logging
underInstalled Namespace
. Leave all other defaults and then clickInstall
.
-
-
Verify the operator installations:
-
Switch to the
Operators
→Installed Operators
page. -
Make sure the
openshift-logging
project is selected. -
In the Status column you should see green checks with either
InstallSucceeded
orCopied
and the text Up to date.During installation an operator might display a
Failed
status. If the operator then installs with anInstallSucceeded
message, you can safely ignore theFailed
message. Also, if you’re using theConsole
tab, you may or maynot see theStatus
column. When in doubt, visit the console via the link.
-
Now that we have the operators installed, along with the CRDs
, we can now
kick off the logging install by creating a Logging CR
. This will define how
we want to install and configure logging.
-
In the OpenShift Console, switch to the the
Administration
→Custom Resource Definitions
page. -
On the
Custom Resource Definitions
page, search forLogging
in the search field and clickClusterLogging
. -
On the
Custom Resource Definition Overview
page, selectInstances
from theActions
menu.If you see a
404
error, don’t panic. While the operator installation succeeded, the operator itself has not finished installing and theCustomResourceDefinition
may not have been created yet. Wait a few moments and then refresh the page. -
On the
Cluster Loggings
page, clickCreate Cluster Logging
. -
In the
YAML
editor, replace the code with the following:
apiVersion: "logging.openshift.io/v1"
kind: "ClusterLogging"
metadata:
name: "instance"
namespace: "openshift-logging"
spec:
managementState: "Managed"
logStore:
type: "elasticsearch"
elasticsearch:
nodeCount: 3
storage:
storageClassName: gp2
size: 100Gi
redundancyPolicy: "SingleRedundancy"
nodeSelector:
node-role.kubernetes.io/infra: ""
resources:
request:
memory: 4G
visualization:
type: "kibana"
kibana:
replicas: 1
nodeSelector:
node-role.kubernetes.io/infra: ""
curation:
type: "curator"
curator:
schedule: "30 3 * * *"
nodeSelector:
node-role.kubernetes.io/infra: ""
collection:
logs:
type: "fluentd"
fluentd: {}
nodeSelector:
node-role.kubernetes.io/infra: ""
Then click Create
.
Now that Logging has been created, let’s verify that things are working.
-
Switch to the
Workloads
→Pods
page. -
Select the
openshift-logging
project.
You should see pods for cluster logging (the operator itself), Elasticsearch, and Fluentd, and Kibana.
Alternatively, you can verify from the command line by using the following command:
oc get pods -n openshift-logging -w
You should eventually see something like:
NAME READY STATUS RESTARTS AGE cluster-logging-operator-5d4b6f7b99-ksr5s 1/1 Running 0 113s collector-2p5fx 2/2 Running 0 26s collector-7lw5r 2/2 Running 0 42s collector-8stvf 2/2 Running 0 32s collector-b7qs8 2/2 Running 0 27s collector-clfsc 2/2 Running 0 16s collector-f2tzf 2/2 Running 0 31s collector-j6hxp 2/2 Running 0 10s collector-kdvj8 2/2 Running 0 30s collector-q6wck 2/2 Running 0 21s collector-sgndk 2/2 Running 0 17s collector-w5ds9 2/2 Running 0 29s collector-zswpb 2/2 Running 0 34s elasticsearch-cdm-mnc985r3-1-5c45b9bd9f-4nx56 2/2 Running 0 70s elasticsearch-cdm-mnc985r3-2-779989b7bb-z9dpp 1/2 Running 0 69s elasticsearch-cdm-mnc985r3-3-6d754c8cbf-fx8wd 1/2 Running 0 68s kibana-655877db88-njsqq 2/2 Running 0 70s
The Collector Pods are deployed as part of a DaemonSet, which is a mechanism to ensure that specific Pods run on specific Nodes in the cluster at all times:
oc get daemonset -n openshift-logging
You will see something like:
oc get daemonset -n openshift-logging NAME DESIRED CURRENT READY UP-TO-DATE AVAILABLE NODE SELECTOR AGE collector 9 9 9 9 9 kubernetes.io/os=linux 83s
You should expect 1 collector
Pod for every Node in your cluster.
Remember that Masters are still Nodes and collector
will run there, too,
to slurp the various logs.
You will also see the storage for ElasticSearch has been automatically provisioned. If you query the PersistentVolumeClaim objects in this project you will see the new storage.
oc get pvc -n openshift-logging
You will see something like:
NAME STATUS VOLUME CAPACITY ACCESS
MODES STORAGECLASS AGE
elasticsearch-elasticsearch-cdm-ggzilasv-1 Bound pvc-f3239564-389c-11ea-bab2-06ca7918708a 100Gi RWO
ocs-storagecluster-ceph-rbd 15m
elasticsearch-elasticsearch-cdm-ggzilasv-2 Bound pvc-f324a252-389c-11ea-bab2-06ca7918708a 100Gi RWO
ocs-storagecluster-ceph-rbd 15m
elasticsearch-elasticsearch-cdm-ggzilasv-3 Bound pvc-f326aa7d-389c-11ea-bab2-06ca7918708a 100Gi RWO
ocs-storagecluster-ceph-rbd 15m