Open Shift Cluster Report

This playbook generates a cluster overview report for an Open Shift 3 Cluster.

Features

Requires no additional dependency
Generates a stand-alone HTML file
Light-weight and fast
Very few oc commands are run on the cluster, data is processed on the ansible node.
Modular design, new report components can be easily added

If you would like to request a new component/features in the report, please open a new github issue.

Usage

Install the required python-jmespath package:

sudo yum install python-jmespath -y

Clone/copy this repository on the ansible node of the cluster:

git clone https://github.com/kxr/ocpcr.git

Change into the directory:

cd ocpcr

And run the cluster-report.yaml playbook

ansible-playbook cluster-report.yaml

If the ansible inventory of the cluster is in a non-default location, specify it with -i:

ansible-playbook cluster-report.yaml -i /root/ocp311-inventory

Once the playbook finishes, it will generate the report in a HTML file OpenShift_Cluster_OverviewReport\.html

The playbook can email the report to one or more recipents as well. Set the email_report: to True and provide the SMTP configuration in cluster-report.yaml

    email_report: True
    email_from: cluster-report@email.com (OpenShift Cluster Overview)
    email_recipients:
      - khizer@email.com
      - someone@email.com
    smtp_host: smtp.gmail.com
    smtp_port: 587
    smtp_user: smtp_auth_user@email.com
    smtp_pass: xxxxxx

Sample

OpenShift Cluster Overview Report 2020-03-09

TODO (WISH LIST)

Check for orphaned pods
Compare the master-config.yaml
Elasticsearch status, indexes
Warning events in the core namespaces
Quick certificate check
Gluster stroage check
Pending CSR check
Core pods (registry, es, prometheus etc.) using persistent storage check
Latency check of core services (api, registry, router) from all nodes
All pods on each node using hostpath/localvolumes
Persistent volumes check
Service serving certificate check
Disk usage of important directories (/var, /var/lib/docker /var/lib/etcd etc.)
Top 10 namespaces using CPU/Memory requests/limits
Handle CRI-O
Show network info i.e, host IPs, oc-hostsubnets
DNS on nodes/pods check
Proxy on nodes/pods check
Users/serviceaccounts with admin/cluster-admin privileges
Imagestream check (deployments using old istreams)
OVS flow, bridges, interfaces check
Create proper document/wiki
Use bootstrap's datatables to allow for interactive table filtering.

Contribution

This is work in progress and I will be adding more components to the report. To understand the architecture of the playbook / jinja templating, please see the wiki section.

Please feel free to write new components and send a pull request.

kxr / ocpcr