$ oc adm new-project logging --node-selector="" $ oc project logging
As an OpenShift Container Platform cluster administrator, you can deploy the EFK stack to aggregate logs for a range of OpenShift Container Platform services. Application developers can view the logs of the projects for which they have view access. The EFK stack aggregates logs from hosts and applications, whether coming from multiple containers or even deleted pods.
The EFK stack is a modified version of the ELK stack and is comprised of:
Elasticsearch: An object store where all logs are stored.
Fluentd: Gathers logs from nodes and feeds them to Elasticsearch.
Kibana: A web UI for Elasticsearch.
After deployment in a cluster, the stack aggregates logs from all nodes and projects into Elasticsearch, and provides a Kibana UI to view any logs. Cluster administrators can view all logs, but application developers can only view logs for projects they have permission to view. The stack components communicate securely.
Managing
Docker Container Logs discusses the use of Aggregated logging is only supported using the |
An Ansible playbook is available to deploy and upgrade aggregated logging. You should familiarize yourself with the advanced installation and configuration section. This provides information for preparing to use Ansible and includes information about configuration. Parameters are added to the Ansible inventory file to configure various areas of the EFK stack.
Review the sizing guidelines to determine how best to configure your deployment.
Ensure that you have deployed a router for the cluster.
Ensure that you have the necessary storage for Elasticsearch. Note that each Elasticsearch replica requires its own storage volume. See Elasticsearch for more information.
Determine if you need highly-available Elasticsearch. A highly-available environment requires
multiple replicas of each shard. By default, OpenShift Container Platform creates one shard for each index and
zero replicas of those shards. To create high availability, set the openshift_logging_es_number_of_replicas
Ansible variable
to a value higher than 1
. High availability also requires at least three Elasticsearch nodes,
each on a different host. See Elasticsearch for more information.
Choose a project. Once deployed, the EFK stack collects logs for every
project within your OpenShift Container Platform cluster. The examples in this section use the
default project logging. The Ansible playbook creates the project for you
if it does not already exist. You will only need to create a project if you want
to specify a node-selector on it. Otherwise, the openshift-logging
role will
create a project.
$ oc adm new-project logging --node-selector="" $ oc project logging
Specifying an empty node selector on the project is recommended, as Fluentd should be deployed throughout the cluster and any selector would restrict where it is deployed. To control component placement, specify node selectors per component to be applied to their deployment configurations. |
Parameters for the EFK deployment may be specified to the inventory host file to override the default parameter values. Read the Elasticsearch and the Fluentd sections before choosing parameters:
By default the Elasticsearch service uses port 9300 for TCP communication between nodes in a cluster. |
Parameter | Description |
---|---|
|
The prefix for logging component images. For example, setting the prefix to registry.access.redhat.com/openshift3/ creates registry.access.redhat.com/openshift3/logging-fluentd:latest. |
|
The version for logging component images. For example, setting the version to v3.5 creates registry.access.redhat.com/openshift3/logging-fluentd:v3.5. |
|
If set to |
|
The URL for the Kubernetes master, this does not need to be public facing but should be accessible from within the cluster. |
|
The public facing URL for the Kubernetes master. This is used for Authentication redirection by the Kibana proxy. |
|
The namespace where Aggregated Logging will be deployed. |
|
Set to |
|
Specify the name of an existing pull secret to be used for pulling component images from an authenticated registry. |
|
The default minimum age (in days) Curator uses for deleting log records. |
|
The hour of the day Curator will run. |
|
The minute of the hour Curator will run. |
|
The timezone Curator uses for figuring out its run time. Provide the
timezone as a string in the tzselect(8) or timedatectl(1) "Region/Locality"
format, for example |
|
The script log level for Curator. |
|
The log level for the Curator process. |
|
The amount of CPU to allocate to Curator. |
|
The amount of memory to allocate to Curator. |
|
A node selector that specifies which nodes are eligible targets for deploying Curator instances. |
|
Equivalent to |
|
Equivalent to |
|
The external host name for web clients to reach Kibana. |
|
The amount of CPU to allocate to Kibana. |
|
The amount of memory to allocate to Kibana. |
|
When |
|
The amount of CPU to allocate to Kibana proxy. |
|
The amount of memory to allocate to Kibana proxy. |
|
The number of replicas to which Kibana should be scaled up. |
|
A node selector that specifies which nodes are eligible targets for deploying Kibana instances. |
|
The public facing key to use when creating the kibana route. |
|
The cert that matches the key when creating the kibana route. |
|
Optional. The CA to goes with the key and cert used when creating the kibana route. |
|
Equivalent to |
|
Equivalent to |
|
Equivalent to |
|
Equivalent to |
|
Equivalent to |
|
Equivalent to |
|
Equivalent to |
|
A node selector that specifies which nodes are eligible targets for deploying Fluentd instances. Any node where Fluentd should run (typically, all) must have this label before Fluentd will be able to run and collect logs. When scaling up the Aggregated Logging cluster after installation,
the As part of the installation, it is recommended that you add the Fluentd node selector label to the list of persisted node labels. |
|
The CPU limit for Fluentd pods. |
|
The memory limit for Fluentd pods. |
|
Set to |
|
Set to |
|
List of nodes that should be labeled for Fluentd to be deployed. |
|
The name of the Elasticsearch service where Fluentd should send logs. |
|
The port for the Elasticsearch service where Fluentd should send logs. |
|
The location of the CA Fluentd uses to communicate with |
|
The location of the client certificate Fluentd uses for |
|
The location of the client key Fluentd uses for |
|
Elasticsearch nodes to deploy. High availability requires at least three or more. |
|
The amount of CPU limit for the Elasticsearch cluster. |
|
Amount of RAM to reserve per Elasticsearch instance. It must be at least 512M. Possible suffixes are G,g,M,m. |
|
The number of replicas per primary shard for each new index. Defaults to '0'. A minimum of |
|
The number of primary shards for every new index created in ES. Defaults to '1'. |
|
A key/value map added to a PVC in order to select specific PVs. |
|
If available for your
cluster, set to |
|
Size of the persistent volume claim to create per ElasticSearch instance. For example, 100G. If omitted, no PVCs are created and ephemeral volumes are used instead. |
|
Prefix for the names of persistent volume claims to be used as storage for
Elasticsearch nodes. A number is appended per node, such as
logging-es-1. If they do not already exist, they are created with size
|
|
The amount of time Elasticsearch will wait before it tries to recover. |
|
Number of a supplemental group ID for access to Elasticsearch storage volumes. Backing volumes should allow access by this group ID. |
|
A node selector specified as a map that determines which nodes are eligible targets
for deploying Elasticsearch nodes. Use this map to place these instances on nodes that are reserved or optimized for running them.
For example, the selector could be |
|
Set to |
|
Equivalent to |
|
Equivalent to |
|
Equivalent to |
|
Equivalent to |
|
Equivalent to |
|
Equivalent to |
|
Equivalent to |
|
Equivalent to |
|
Equivalent to |
|
Equivalent to |
|
Equivalent to |
|
Equivalent to |
|
Equivalent to |
|
Equivalent to |
|
A node selector that specifies which nodes are eligible targets
for deploying Elasticsearch nodes. This can be used to place
these instances on nodes reserved or optimized for running them.
For example, the selector could be |
|
A node selector that specifies which nodes are eligible targets for deploying Kibana instances. |
|
A node selector that specifies which nodes are eligible targets for deploying Curator instances. |
The EFK stack is deployed using an Ansible playbook to the EFK components. Run the playbook from the default OpenShift Ansible location using the default inventory file.
$ ansible-playbook [-i </path/to/inventory>] \ /usr/share/ansible/openshift-ansible/playbooks/byo/openshift-cluster/openshift-logging.yml
Running the playbook deploys all resources needed to support the stack; such as Secrets, ServiceAccounts, and DeploymentConfigs. The playbook waits to deploy the component pods until the stack is running. If the wait steps fail, the deployment could still be successful; it may be retrieving the component images from the registry which can take up to a few minutes. You can watch the process with:
$ oc get pods -w
They will eventually enter Running status. For additional details about the status of the pods during deployment by retrieving associated events:
$ oc describe pods/<pod_name>
Check the logs if the pods do not run successfully:
$ oc logs -f <pod_name>
This section describes adjustments that you can make to deployed components.
The logs for the default, openshift, and openshift-infra projects are automatically aggregated and grouped into the .operations item in the Kibana interface. The project where you have deployed the EFK stack (logging, as documented here) is not aggregated into .operations and is found under its ID. |
If you set openshift_logging_use_ops
to true in your inventory file, Fluentd is
configured to split logs between the main Elasticsearch cluster and another
cluster reserved for operations logs, which are defined as node system logs and
the projects default, openshift, and openshift-infra. Therefore, a
separate Elasticsearch cluster, a separate Kibana, and a separate Curator are
deployed to index, access, and manage operations logs. These deployments are set
apart with names that include -ops
. Keep these separate deployments in mind if
you enable this option. Most of the following discussion also applies to the
operations cluster if present, just with the names changed to include -ops
.
Elasticsearch (ES) is an object store where all logs are stored.
Elasticsearch organizes the log data into datastores, each called an index. Elasticsearch subdivides each index into multiple pieces called shards, which it spreads across a set of Elasticsearch nodes in your cluster. You can configure Elasticsearch to make copies of the shards, called replicas. Elasticsearch also spreads replicas across the Elactisearch nodes. The combination of shards and replicas is intended to provide redundancy and resilience to failure. For example, if you configure three shards for the index with one replica, Elasticsearch generates a total of six shards for that index: three primary shards and three replicas as a backup.
The OpenShift Container Platform logging installer ensures each Elasticsearch node is deployed using a unique deployment configuration that includes its own storage volume.
You can create an additional deployment configuration for each Elasticsearch node you add to the logging system.
During installation, you can use the openshift_logging_es_cluster_size
Ansible variable to specify the number of Elasticsearch nodes.
Alternatively, you can scale up your existing cluster by modifying by modifying the
openshift_logging_es_cluster_size
in the inventory file and re-running the
logging playbook. Additional clustering parameters can be modified and are
described in Specifying Logging Ansible Variables.
Refer to Elastic’s documentation for considerations involved in choosing storage and network location as directed below.
A highly-available Elasticsearch environment requires at least three Elasticsearch nodes,
each on a different host, and setting the |
Viewing all Elasticsearch Deployments
To view all current Elasticsearch deployments:
$ oc get dc --selector logging-infra=elasticsearch
Configuring Elasticsearch for High Availability
A highly-available Elasticsearch environment requires at least three Elasticsearch nodes,
each on a different host, and setting the openshift_logging_es_number_of_replicas
Ansible variable
to a value of 1
or higher to create replicas.
Use the following scenarios as a guide for an OpenShift Container Platform cluster with three Elasticsearch nodes:
If you can tolerate one Elasticsearch node going down,
set openshift_logging_es_number_of_replicas
to 1
. This ensures
that two nodes have a copy of all of the Elasticsearch data in the cluster.
If you must tolerate two Elasticsearch nodes going down,
set openshift_logging_es_number_of_replicas
to 2
. This ensures that
every node has a copy of all of the Elasticsearch data in the cluster.
Note that there is a trade-off between high availability and performance.
For example, having openshift_logging_es_number_of_replicas=2
and
openshift_logging_es_number_of_shards=3
requires Elasticsearch to spend
significant resources replicating the shard data among the nodes in the cluster.
Also, using a higher number of replicas requires doubling or tripling the data storage
requirements on each node, so you must take that into account when planning
persistent storage for Elasticsearch.
Considerations when Configuring the Number of Shards
For the openshift_logging_es_number_of_shards
parameter, consider:
For higher performance, increase the number of shards. For example, in a three
node cluster, set openshift_logging_es_number_of_shards=3
. This will cause
each index to be split into three parts (shards), and the load for processing the
index will be spread out over all 3 nodes.
If you have a large number of projects, you might see performance degradation if you have more than a few thousand shards in the cluster. Either reduce the number of shards or reduce the curation time.
If you have a small number of very large indices, you might want to configure
openshift_logging_es_number_of_shards=3
or higher. Elasticsearch recommends
using a maximum shard size of less than 50 GB.
Node Selector
Because Elasticsearch can use a lot of resources, all members of a cluster should have low latency network connections to each other and to any remote storage. Ensure this by directing the instances to dedicated nodes, or a dedicated region within your cluster, using a node selector.
To configure a node selector, specify the openshift_logging_es_nodeselector
configuration option in the inventory file. This applies to all Elasticsearch
deployments; if you need to individualize the node selectors, you must manually
edit each deployment configuration after deployment. The node selector is
specified as a python compatible dict. For example, {"node-type":"infra",
"region":"east"}
.
Persistent Elasticsearch Storage
By default, the openshift_logging
Ansible role creates an ephemeral
deployment in which all of a pod’s data is lost upon restart. For production
usage, specify a persistent storage volume for each Elasticsearch deployment
configuration. You can create the necessary
persistent
volume claims before deploying or have them created for you. The PVCs must be
named to match the openshift_logging_es_pvc_prefix
setting, which defaults to
logging-es-
; each PVC name will have a sequence number added to it, so
logging-es-1
, logging-es-2
, and so on. If a PVC needed for the deployment
exists already, it is used; if not, and openshift_logging_es_pvc_size
has been
specified, it is created with a request for that size.
Using NFS storage as a volume or a persistent volume, or using NAS such as Gluster, is not supported for Elasticsearch storage, as Lucene relies on file system behavior that NFS does not supply. Data corruption and other problems can occur. If NFS storage is a requirement, you can allocate a large file on a volume to serve as a storage device and mount it locally on one host. For example, if your NFS storage volume is mounted at /nfs/storage: $ truncate -s 1T /nfs/storage/elasticsearch-1 $ mkfs.xfs /nfs/storage/elasticsearch-1 $ mount -o loop /nfs/storage/elasticsearch-1 /usr/local/es-storage $ chown 1000:1000 /usr/local/es-storage Then, use /usr/local/es-storage as a host-mount as described below. Use a different backing file as storage for each Elasticsearch replica. This loopback must be maintained manually outside of OpenShift Container Platform, on the node. You must not maintain it from inside a container. |
It is possible to use a local disk volume (if available) on each node host as storage for an Elasticsearch replica. Doing so requires some preparation as follows.
The relevant service account must be given the privilege to mount and edit a local volume:
$ oc adm policy add-scc-to-user privileged \ system:serviceaccount:logging:aggregated-logging-elasticsearch (1)
1 | Use the project you created earlier (for example, logging) when running the logging playbook. |
Each Elasticsearch replica definition must be patched to claim that privilege, for example (change to --selector component=es-ops
for Ops cluster):
$ for dc in $(oc get deploymentconfig --selector component=es -o name); do oc scale $dc --replicas=0 oc patch $dc \ -p '{"spec":{"template":{"spec":{"containers":[{"name":"elasticsearch","securityContext":{"privileged": true}}]}}}}' done
The Elasticsearch replicas must be located on the correct nodes to use the local storage, and should not move around even if those nodes are taken down for a period of time. This requires giving each Elasticsearch replica a node selector that is unique to a node where an administrator has allocated storage for it. To configure a node selector, edit each Elasticsearch deployment configuration and add or edit the nodeSelector section to specify a unique label that you have applied for each desired node:
apiVersion: v1 kind: DeploymentConfig spec: template: spec: nodeSelector: logging-es-node: "1" (1)
1 | This label should uniquely identify a replica with a single node that bears that
label, in this case logging-es-node=1 . Use the oc label command to apply
labels to nodes as needed. |
To automate applying the node selector you can instead use the oc patch
command:
$ oc patch dc/logging-es-<suffix> \ -p '{"spec":{"template":{"spec":{"nodeSelector":{"logging-es-node":"1"}}}}}'
Once these steps are taken, a local host mount can be applied to each replica as in this example
(where we assume storage is mounted at the same path on each node) (change to --selector component=es-ops
for Ops cluster):
$ for dc in $(oc get deploymentconfig --selector component=es -o name); do oc set volume $dc \ --add --overwrite --name=elasticsearch-storage \ --type=hostPath --path=/usr/local/es-storage oc rollout latest $dc oc scale $dc --replicas=1 done
Changing the Scale of Elasticsearch
If you need to scale up the number of Elasticsearch nodes in your cluster, you can create a deployment configuration for each Elasticsearch node you want to add.
Due to the nature of persistent volumes and how Elasticsearch is configured to store its data and recover the cluster, you cannot simply increase the replicas in an Elasticsearch deployment configuration.
The simplest way to change the scale of Elasticsearch is to modify the inventory host file a re-run the logging playbook as desribed previously. Assuming you have supplied persistent storage for the deployment, this should not be disruptive.
If you do not wish to reinstall, for instance because you have made customizations that you would like to preserve, then it is possible to add new Elasticsearch deployment configurations to the cluster using a template supplied by the deployer. This requires a more complicated procedure however.
Allowing cluster-reader to view operations logs
By default, only cluster-admin
users are granted access in Elasticsearch and
Kibana to view operations logs. To allow cluster-reader
users to also view these
logs, update the value of openshift.operations.allow_cluster_reader
in the
Elasticsearch configmap to true
:
$ oc edit configmap/logging-elasticsearch
Please note that changes to the configmap might not appear until after redeploying
the pods. Persisting these changes across deployments can be accomplished by setting
openshift_logging_es_allows_cluster_reader
to true
in the inventory file.
Fluentd is deployed as a DaemonSet that deploys replicas according to a node
label selector, which you can specify with the inventory parameter
openshift_logging_fluentd_nodeselector
and the default is logging-infra-fluentd
.
As part of the OpenShift cluster installation, it is recommended that you add the
Fluentd node selector to the list of persisted
node labels.
Having Fluentd Use the Systemd Journal as the Log Source
By default, Fluentd reads from /var/log/messages and /var/log/containers/<container>.log for system logs and container logs, respectively. You can instead use the systemd journal as the log source. There are three inventory parameters available:
Parameter | Description |
---|---|
|
The default is empty, which configures Fluentd to check which log driver Docker
is using. If Docker is using Aggregated logging is only supported using the |
|
If this setting is |
As of OpenShift Container Platform 3.3, Fluentd no longer reads historical log files when using the JSON file log driver. In situations where clusters have a large number of log files and are older than the EFK deployment, this avoids delays when pushing the most recent logs into Elasticsearch. Curator deleting logs are migrated soon after they are added to Elasticsearch. |
It may require several minutes, or hours, depending on the size of your
journal, before any new log entries are available in Elasticsearch, when using
|
It is highly recommended that you use the default value for Aggregated logging is only supported using the |
Having Fluentd Send Logs to Another Elasticsearch
The use of |
You can configure Fluentd to send a copy of each log message to both the Elasticsearch instance included with OpenShift Container Platform aggregated logging, and to an external Elasticsearch instance. For example, if you already have an Elasticsearch instance set up for auditing purposes, or data warehousing, you can send a copy of each log message to that Elasticsearch.
This feature is controlled via environment variables on Fluentd, which can be modified as described below.
If its environment variable ES_COPY
is true, Fluentd sends a copy of the
logs to another Elasticsearch. The names for the copy variables are just like
the current ES_HOST
, OPS_HOST
, and other variables, except that they add
_COPY
: ES_COPY_HOST
, OPS_COPY_HOST
, and so on. There are some
additional parameters added:
ES_COPY_SCHEME
, OPS_COPY_SCHEME
- can use either http
or https
- defaults
to https
ES_COPY_USERNAME
, OPS_COPY_USERNAME
- user name to use to authenticate to
Elasticsearch using username/password auth
ES_COPY_PASSWORD
, OPS_COPY_PASSWORD
- password to use to authenticate to
Elasticsearch using username/password auth
Sending logs directly to an AWS Elasticsearch instance is not supported. Use
Fluentd Secure Forward to direct logs to
an instance of Fluentd that you control and that is configured with the
|
To set the parameters:
Edit the DaemonSet for Fluentd:
$ oc edit -n logging ds logging-fluentd
Add or edit the environment variable ES_COPY
to have the value "true"
(with the quotes),
and add or edit the COPY variables listed above.
These changes will not be persisted across multiple runs of the logging playbook. You will need to edit the DaemonSet each time to update environment variables. |
Configuring Fluentd to Send Logs to an External Log Aggregator
You can configure Fluentd to send a copy of its logs to an external log
aggregator, and not the default Elasticsearch, using the secure-forward
plug-in. From there, you can further process log records after the locally
hosted Fluentd has processed them.
The logging deployment provides a secure-forward.conf
section in the Fluentd configmap
for configuring the external aggregator:
<store> @type secure_forward self_hostname pod-${HOSTNAME} shared_key thisisasharedkey secure yes enable_strict_verification yes ca_cert_path /etc/fluent/keys/your_ca_cert ca_private_key_path /etc/fluent/keys/your_private_key ca_private_key_passphrase passphrase <server> host ose1.example.com port 24284 </server> <server> host ose2.example.com port 24284 standby </server> <server> host ose3.example.com port 24284 standby </server> </store>
This can be updated using the oc edit
command:
$ oc edit configmap/logging-fluentd
Certificates to be used in secure-forward.conf
can be added to the existing
secret that is mounted on the Fluentd pods. The your_ca_cert
and
your_private_key
values must match what is specified in secure-forward.conf
in configmap/logging-fluentd
:
$ oc patch secrets/logging-fluentd --type=json \ --patch "[{'op':'add','path':'/data/your_ca_cert','value':'$(base64 /path/to/your_ca_cert.pem)'}]" $ oc patch secrets/logging-fluentd --type=json \ --patch "[{'op':'add','path':'/data/your_private_key','value':'$(base64 /path/to/your_private_key.pem)'}]"
Replace |
When configuring the external aggregator, it must be able to accept messages securely from Fluentd.
If the external aggregator is another Fluentd server, it must have the
fluent-plugin-secure-forward
plug-in installed and make use of the input
plug-in it provides:
<source> @type secure_forward self_hostname ${HOSTNAME} bind 0.0.0.0 port 24284 shared_key thisisasharedkey secure yes cert_path /path/for/certificate/cert.pem private_key_path /path/for/certificate/key.pem private_key_passphrase secret_foo_bar_baz </source>
Further explanation of how to set up the fluent-plugin-secure-forward
plug-in
can be found
here.
Throttling logs in Fluentd
For projects that are especially verbose, an administrator can throttle down the rate at which the logs are read in by Fluentd before being processed.
Throttling can contribute to log aggregation falling behind for the configured projects; log entries can be lost if a pod is deleted before Fluentd catches up. |
Throttling does not work when using the systemd journal as the log source. The throttling implementation depends on being able to throttle the reading of the individual log files for each project. When reading from the journal, there is only a single log source, no log files, so no file-based throttling is available. There is not a method of restricting the log entries that are read into the Fluentd process. |
To tell Fluentd which projects it should be restricting, edit the throttle configuration in its ConfigMap after deployment:
$ oc edit configmap/logging-fluentd
The format of the throttle-config.yaml key is a YAML file that contains project names and the desired rate at which logs are read in on each node. The default is 1000 lines at a time per node. For example:
logging: read_lines_limit: 500 test-project: read_lines_limit: 10 .operations: read_lines_limit: 100
To access the Kibana console from the OpenShift Container Platform web console, add the
loggingPublicURL
parameter in the /etc/origin/master/master-config.yaml
file, with the URL of the Kibana console (the kibana-hostname
parameter).
The value must be an HTTPS URL:
... assetConfig: ... loggingPublicURL: "https://kibana.example.com" ...
Setting the loggingPublicURL
parameter creates a View Archive button on the
OpenShift Container Platform web console under the Browse → Pods → <pod_name> →
Logs tab. This links to the Kibana console.
You can scale the Kibana deployment as usual for redundancy:
$ oc scale dc/logging-kibana --replicas=2
To ensure the scale persists across multiple executions of the logging playbook,
make sure to update the |
You can see the user interface by visiting the site specified by the
openshift_logging_kibana_hostname
variable.
See the Kibana documentation for more information on Kibana.
Curator allows administrators to configure scheduled Elasticsearch maintenance operations to be performed automatically on a per-project basis. It is scheduled to perform actions daily based on its configuration. Only one Curator pod is recommended per Elasticsearch cluster. Curator is configured via a YAML configuration file with the following structure:
$PROJECT_NAME: $ACTION: $UNIT: $VALUE $PROJECT_NAME: $ACTION: $UNIT: $VALUE ...
The available parameters are:
Variable Name | Description |
---|---|
|
The actual name of a project, such as myapp-devel. For OpenShift Container Platform operations
logs, use the name |
|
The action to take, currently only |
|
One of |
|
An integer for the number of units. |
|
Use |
|
(Number) the hour of the day in 24-hour format at which to run the Curator jobs. For
use with |
|
(Number) the minute of the hour at which to run the Curator jobs. For use with |
For example, to configure Curator to:
delete indices in the myapp-dev project older than 1 day
delete indices in the myapp-qe project older than 1 week
delete operations logs older than 8 weeks
delete all other projects indices after they are 30 days
old
run the Curator jobs at midnight every day
Use:
myapp-dev: delete: days: 1 myapp-qe: delete: weeks: 1 .operations: delete: weeks: 8 .defaults: delete: days: 30 runhour: 0 runminute: 0
When you use |
The openshift_logging
Ansible role provides a ConfigMap from which Curator reads its
configuration. You may edit or replace this ConfigMap to reconfigure
Curator. Currently the logging-curator
ConfigMap is used to
configure both your ops and non-ops Curator instances. Any .operations
configurations will be in the same location as your application logs
configurations.
To edit the provided ConfigMap to configure your Curator instances:
$ oc edit configmap/logging-curator
To replace the provided ConfigMap instead:
$ create /path/to/mycuratorconfig.yaml $ oc create configmap logging-curator -o yaml \ --from-file=config.yaml=/path/to/mycuratorconfig.yaml | \ oc replace -f -
After you make your changes, redeploy Curator:
$ oc rollout latest dc/logging-curator $ oc rollout latest dc/logging-curator-ops
Remove everything generated during the deployment.
$ ansible-playbook [-i </path/to/inventory>] \ /usr/share/ansible/openshift-ansible/playbooks/byo/openshift-cluster/openshift-logging.yml \ -e openshift_logging_install_logging=False
Using the Kibana console with OpenShift Container Platform can cause problems that are easily solved, but are not accompanied with useful error messages. Check the following troubleshooting sections if you are experiencing any problems when deploying Kibana on OpenShift Container Platform:
Login Loop
The OAuth2 proxy on the Kibana console must share a secret with the master host’s OAuth2 server. If the secret is not identical on both servers, it can cause a login loop where you are continuously redirected back to the Kibana login page.
To fix this issue, delete the current OAuthClient, and use openshift-ansible
to re-run the openshift_logging
role:
$ oc delete oauthclient/kibana-proxy $ ansible-playbook [-i </path/to/inventory>] \ /usr/share/ansible/openshift-ansible/playbooks/byo/openshift-cluster/openshift-logging.yml
Cryptic Error When Viewing the Console
When attempting to visit the Kibana console, you may receive a browser error instead:
{"error":"invalid_request","error_description":"The request is missing a required parameter, includes an invalid parameter value, includes a parameter more than once, or is otherwise malformed."}
This can be caused by a mismatch between the OAuth2 client and server. The return address for the client must be in a whitelist so the server can securely redirect back after logging in.
Fix this issue by replacing the OAuthClient entry:
$ oc delete oauthclient/kibana-proxy $ ansible-playbook [-i </path/to/inventory>] \ /usr/share/ansible/openshift-ansible/playbooks/byo/openshift-cluster/openshift-logging.yml
If the problem persists, check that you are accessing Kibana at a URL listed in the OAuth client. This issue can be caused by accessing the URL at a forwarded port, such as 1443 instead of the standard 443 HTTPS port. You can adjust the server whitelist by editing the OAuth client:
$ oc edit oauthclient/kibana-proxy
503 Error When Viewing the Console
If you receive a proxy error when viewing the Kibana console, it could be caused by one of two issues.
First, Kibana may not be recognizing pods. If Elasticsearch is slow in starting up, Kibana may timeout trying to reach it. Check whether the relevant service has any endpoints:
$ oc describe service logging-kibana Name: logging-kibana [...] Endpoints: <none>
If any Kibana pods are live, endpoints will be listed. If they are not, check the state of the Kibana pods and deployment. You may need to scale the deployment down and back up again.
The second possible issue may be caused if the route for accessing the Kibana service is masked. This can happen if you perform a test deployment in one project, then deploy in a different project without completely removing the first deployment. When multiple routes are sent to the same destination, the default router will only route to the first created. Check the problematic route to see if it is defined in multiple places:
$ oc get route --all-namespaces --selector logging-infra=support
F-5 Load Balancer and X-Forwarded-For Enabled
If you are attempting to use a F-5 load balancer in front of Kibana with
X-Forwarded-For
enabled, this can cause an issue in which the Elasticsearch
Searchguard
plug-in is unable to correctly accept connections from Kibana.
Kibana: Unknown error while connecting to Elasticsearch Error: Unknown error while connecting to Elasticsearch Error: UnknownHostException[No trusted proxies]
To configure Searchguard to ignore the extra header:
Scale down all Fluentd pods.
Scale down Elasticsearch after the Fluentd pods have terminated.
Add searchguard.http.xforwardedfor.header: DUMMY
to the Elasticsearch
configuration section.
$ oc edit configmap/logging-elasticsearch (1)
1 | This approach requires that Elasticsearch’s configurations are within a ConfigMap. |
Scale Elasticsearch back up.
Scale up all Fluentd pods.
Fluentd sends logs to the value of the ES_HOST
, ES_PORT
, OPS_HOST
,
and OPS_PORT
environment variables of the Elasticsearch deployment
configuration. The application logs are directed to the ES_HOST
destination,
and operations logs to OPS_HOST
.
Sending logs directly to an AWS Elasticsearch instance is not supported. Use
Fluentd Secure Forward to direct logs to
an instance of Fluentd that you control and that is configured with the
|
To direct logs to a specific Elasticsearch instance, edit the deployment configuration and replace the value of the above variables with the desired instance:
$ oc edit dc/<deployment_configuration>
For an external Elasticsearch instance to contain both application and
operations logs, you can set ES_HOST
and OPS_HOST
to the same destination,
while ensuring that ES_PORT
and OPS_PORT
also have the same value.
If your externally hosted Elasticsearch instance does not use TLS, update the
_CLIENT_CERT
, _CLIENT_KEY
, and _CA
variables to be empty. If it does
use TLS, but not mutual TLS, update the _CLIENT_CERT
and _CLIENT_KEY
variables to be empty and patch or recreate the logging-fluentd secret with
the appropriate _CA
value for communicating with your Elasticsearch instance.
If it uses Mutual TLS as the provided Elasticsearch instance does, patch or
recreate the logging-fluentd secret with your client key, client cert, and CA.
If you are not using the provided Kibana and Elasticsearch images, you will not have the same multi-tenant capabilities and your data will not be restricted by user access to a particular project. |
As of logging version 3.2.0, an administrator certificate, key, and CA that can be used to communicate with and perform administrative operations on Elasticsearch are provided within the logging-elasticsearch secret.
To confirm whether or not your EFK installation provides these, run: $ oc describe secret logging-elasticsearch |
If they are not available, refer to Manual Upgrades to ensure you are on the latest version first.
Connect to an Elasticsearch pod that is in the cluster on which you are attempting to perform maintenance.
To find a pod in a cluster use either:
$ oc get pods -l component=es -o name | head -1 $ oc get pods -l component=es-ops -o name | head -1
Connect to a pod:
$ oc rsh <your_Elasticsearch_pod>
Once connected to an Elasticsearch container, you can use the certificates mounted from the secret to communicate with Elasticsearch per its Indices APIs documentation.
Fluentd sends its logs to Elasticsearch using the index format project.{project_name}.{project_uuid}.YYYY.MM.DD where YYYY.MM.DD is the date of the log record.
For example, to delete all logs for the logging project with uuid 3b3594fa-2ccd-11e6-acb7-0eb6b35eaee3 from June 15, 2016, we can run:
$ curl --key /etc/elasticsearch/secret/admin-key \ --cert /etc/elasticsearch/secret/admin-cert \ --cacert /etc/elasticsearch/secret/admin-ca -XDELETE \ "https://localhost:9200/project.logging.3b3594fa-2ccd-11e6-acb7-0eb6b35eaee3.2016.06.15"
If the Docker log driver has changed from json-file
to journald
and Fluentd
was previously configured with USE_JOURNAL=False
, then it will not be able to
pick up any new logs that are created. When the Fluentd daemonset is configured
with the default value for USE_JOURNAL
, then it will detect the Docker log
driver upon pod start-up, and configure itself to pull from the appropriate source.
To update Fluentd to detect the correct source upon start-up:
Remove the label from nodes where Fluentd is deployed:
$ oc label node --all logging-infra-fluentd- (1)
1 | This example assumes use of the default Fluentd node selector and it being deployed on all nodes. |
Update the daemonset/logging-fluentd
USE_JOURNAL
value to be empty:
$ oc patch daemonset/logging-fluentd \ -p '{"spec":{"template":{"spec":{"containers":[{"name":"fluentd-elasticsearch","env":[{"name": "USE_JOURNAL", "value":""}]}]}}}}'
Relabel your nodes to schedule Fluentd deployments:
$ oc label node --all logging-infra-fluentd=true (1)
1 | This example assumes use of the default Fluentd node selector and it being deployed on all nodes. |