Skip to content

Commit

Permalink
Merge pull request #657 from SUSE/metrics-discrepancies
Browse files Browse the repository at this point in the history 
Update Metrics content to match Github README
  • Loading branch information
@btat
btat authored Jan 3, 2020
2 parents bbc345f + 557bb87 commit 6af049d
Showing 1 changed file with 191 additions and 35 deletions.
226 changes: 191 additions & 35 deletions xml/cap_depl_stratos.xml
View file
Original file line number Diff line number Diff line change
Expand Up @@ -621,47 +621,200 @@ gp2scoped kubernetes.io/aws-ebs 1d
<title>Stratos Metrics</title>

<para>
Stratos can show metrics data from
<link xlink:href="https://prometheus.io/">Prometheus</link> for both &cf;
and &kube;.
Stratos Metrics provides a &helm; chart for deploying Prometheus (see
<link xlink:href="https://prometheus.io/"/>) and the following metrics
exporters to &kube;:
</para>
<itemizedlist>
<listitem>
<para>
&cf; Firehose Exporter (enabled by default)
</para>
</listitem>
<listitem>
<para>
&cf; Exporter (disabled by default)
</para>
</listitem>
<listitem>
<para>
&kube; State Metrics Exporter (disabled by default)
</para>
</listitem>
</itemizedlist>
<para>
The Stratos Metrics &helm; chart deploys a Prometheus server and the
configured Exporters and fronts the Prometheus server with an nginx server
to provide authenticated access to Prometheus (currently basic authentication
over HTTPS).
</para>
<para>
When required by configuration, it also contains an initialization script
that will setup users in the UAA that have correct scopes/permissions to be
able to read data from the &cf; Firehose and/or API.
</para>
<para>
Lastly, the &helm; chart generates a small metadata file in the root of the
nginx server that is used by Stratos to determine which &cf; and &kube;
clusters the Prometheus server is providing Metrics for.
</para>
<para>
To learn more about Stratos Metrics and its full list of configuration
options, see <link xlink:href="https://github.com/SUSE/stratos-metrics"/>.
</para>

<sect2 xml:id="sec-cap-stratos-metrics-exporter-config">
<title>Exporter Configuration</title>
<sect3 xml:id="sec-cap-stratos-metrics-exporter-config-firehose">
<title>Firehose Exporter</title>

<para>
This exporter can be enabled/disabled via the &helm; value
<literal>firehoseExporter.enabled</literal>. By default this exporter is
enabled.
</para>
<para>
You must provide the following Helm chart values for this Exporter to work correctly:
</para>
<itemizedlist>
<listitem>
<para>
<literal>cloudFoundry.apiEndpoint</literal> - API Endpoint of the &cf;
API Server
</para>
</listitem>
<listitem>
<para>
<literal>cloudFoundry.uaaAdminClient</literal> - Admin client of the UAA
used by the &cf; server
</para>
</listitem>
<listitem>
<para>
<literal>cloudFoundry.uaaAdminClientSecret</literal> - Admin client
secret of the UAA used by the &cf; server
</para>
</listitem>
<listitem>
<para>
<literal>cloudFoundry.skipSslVerification</literal> - Whether to skip SSL
verification when communicating with &cf; and the UAA APIs
</para>
</listitem>
</itemizedlist>
<para>
You can scale the firehose nozzle in Stratos Metrics by specifying the
following override:
</para>
<screen>
firehoseExporter:
instances: 1
</screen>
<para>
Please note, the number of firehose nozzles should be proportional to the
number of Traffic Controllers in your &cf; (see docs at
<link xlink:href="https://docs.cloudfoundry.org/loggregator/log-ops-guide.html"/>).
Otherwise, Loggregator will not split the firehose between the nozzles.
</para>
</sect3>
<sect3 xml:id="sec-cap-stratos-metrics-exporter-config-cloud-foundry">
<title>&cf; Exporter</title>

<para>
This exporter can be enabled/disabled via the &helm; value
<literal>cfExporter.enabled</literal>. By default this exporter is
disabled.
</para>
<para>
You must provide the following &helm; chart values for this Exporter to
work correctly:
</para>
<itemizedlist>
<listitem>
<para>
<literal>cloudFoundry.apiEndpoint</literal> - API Endpoint of the &cf;
API Server
</para>
</listitem>
<listitem>
<para>
<literal>cloudFoundry.uaaAdminClient</literal> - Admin client of the UAA
used by the &cf; server
</para>
</listitem>
<listitem>
<para>
<literal>cloudFoundry.uaaAdminClientSecret</literal> - Admin client
secret of the UAA used by the &cf; server
</para>
</listitem>
<listitem>
<para>
<literal>cloudFoundry.skipSslVerification</literal> - Whether to skip SSL
verification when communicating with &cf; and the UAA APIs
</para>
</listitem>
</itemizedlist>
</sect3>

<sect3 xml:id="sec-cap-stratos-metrics-exporter-config-kubernetes">
<title>&kube; Monitoring</title>

<para>
This exporter can be enabled/disabled via the &helm; value
<literal>prometheus.kubeStateMetrics.enabled</literal>. By default this
exporter is disabled.
</para>
<para>
You must provide the following &helm; chart values for this Exporter to
work correctly:
</para>
<itemizedlist>
<listitem>
<para>
<literal>kubernetes.apiEndpoint</literal> - The API Endpoint of the
&kube; API Server
</para>
</listitem>
</itemizedlist>
</sect3>
</sect2>

<sect2 xml:id="sec-cap-stratos-metrics-install">
<title>Install Stratos Metrics with &helm;</title>
<para>
In order to display metrics data with Stratos, you need to deploy the
<literal>stratos-metrics</literal> &helm; chart - this deploys Prometheus
with the necessary exporters that collect data from &cf; and
&kube;. It also wraps Prometheus with an
<link xlink:href="http://nginx.org/">nginx</link> server to provide
authentication.
<literal>stratos-metrics</literal> &helm; chart. As with deploying Stratos,
you should deploy the metrics &helm; chart using the same
<filename>scf-config-values.yaml</filename> file that was used for deploying
<literal>scf</literal> and <literal>uaa</literal>.
</para>
<para>
As with deploying Stratos, you should deploy the metrics &helm; chart using
the same <filename>scf-config-values.yaml</filename> file that was used for
deploying <literal>scf</literal> and <literal>uaa</literal>.
Additionally, create a new YAML file, named
<filename>stratos-metrics-values.yaml</filename> in this example, for
configuration options specific to Stratos Metrics.
</para>
<para>
Create a new yaml file named
<filename>stratos-metrics-values.yaml</filename>, with the following
contents:
The following is an example <filename>stratos-metrics-values.yaml</filename>
file.
</para>
<screen>env:
<screen>cloudFoundry:
apiEndpoint: <replaceable>https://api.example.com</replaceable>
uaaAdminClient: admin
uaaAdminClientSecret: <replaceable>password</replaceable>
skipSslVerification: "true"
env:
DOPPLER_PORT: 443
kubernetes:
apiEndpoint: <replaceable>kube_server_address.example.com</replaceable>
metrics:
username: <replaceable>username</replaceable>
password: <replaceable>password</replaceable>
prometheus:
kubeStateMetrics:
enabled: true
server:
storageClass: <replaceable>"persistent"</replaceable>
nginx:
username: <replaceable>username</replaceable>
password: <replaceable>password</replaceable>
services:
loadbalanced: true</screen>
<para>
Expand All @@ -670,37 +823,40 @@ services:
<itemizedlist>
<listitem>
<para>
<literal>apiEndpoint</literal> is the same URL that you used when
registering your &kube; environment with Stratos (the &kube; API
Server URL)
<literal>kubernetes.apiEndpoint</literal> is the same URL that you used
when registering your &kube; environment with Stratos (the &kube; API
Server URL).
</para>
</listitem>
<listitem>
<para>
<literal>storageClass</literal> is the storage class to be used by Stratos
Metrics. If a storage class is not assigned, the default storage class
will be used. If a storage class is not specified and there is no default
storage class, the <literal>prometheus</literal> pod will fail to start.
<literal>prometheus.server.storageClass</literal> is the storage class to
be used by Stratos Metrics. If a storage class is not assigned, the
default storage class will be used. If a storage class is not specified
and there is no default storage class, the <literal>prometheus</literal>
pod will fail to start.
</para>
</listitem>
<listitem>
<para>
<literal>username</literal> should be chosen by you as the username that
you will use when connecting to Stratos Metrics
<literal>metrics.username</literal> is the username used to authenticate
with the nginx server that fronts Prometheus. This username is also used
during the <xref linkend="sec-cap-stratos-metrics-connect"/>) process.
</para>
</listitem>
<listitem>
<para>
<literal>password</literal> should be chosen by you as the password that
you will use when connecting to Stratos Metrics. Ensure to choose a secure
password
<literal>metrics.password</literal> is the password used to authenticate
with the nginx server that fronts Prometheus. This username is also used
during the <xref linkend="sec-cap-stratos-metrics-connect"/>) process.
Ensure a secure password is chosen.
</para>
</listitem>
<listitem>
<para>
<literal>loadbalanced</literal> is set to <literal>true</literal> if your &kube;
deployment supports automatic configuration of a load balancer (for example,
AKS, EKS, and GKE)
<literal>services.loadbalanced</literal> is set to <literal>true</literal>
if your &kube; deployment supports automatic configuration of a load
balancer (for example, AKS, EKS, and GKE).
</para>
</listitem>
</itemizedlist>
Expand Down Expand Up @@ -830,14 +986,14 @@ susecf-metrics-metrics-nginx NodePort 172.28.107.209 10.86.101.115,172.28.
<step>
<para>
Enter the username for your Stratos Metrics instance. This will be the
<literal>nginx.username</literal> defined in your
<literal>metrics.username</literal> defined in your
<filename>stratos-metrics-values.yaml</filename> file.
</para>
</step>
<step>
<para>
Enter the password for your Stratos Metrics instance. This will be the
<literal>nginx.password</literal> defined in your
<literal>metrics.password</literal> defined in your
<filename>stratos-metrics-values.yaml</filename> file.
</para>
</step>
Expand Down

0 comments on commit 6af049d

Please sign in to comment.