Ask an OpenShift Admin Office Hour - Day 2 Operations, Part 2

2021년 4월 23일Andrew Sullivan2분 읽기

Continuing our closer look at Day 2 Operations, in part two of two today, we reviewed the network and storage configuration tasks to be done after the cluster is deployed. No matter how well and thoroughly the cluster is configured for the application workload, if the application administrators, developers, and users can't access those resources, it's all meaningless! We ran out of time before we could look in-depth at authentication and authorization, but please find the important links below - and we’re scheduling a dedicated session on that topic in the future!

As always, please see the list below for additional links to specific topics, questions, and supporting materials for the episode!

If you’re interested in more streaming content, please subscribe to the OpenShift.tv streaming calendar to see the upcoming episode topics and to receive any schedule changes. If you have questions or topic suggestions for the Ask an OpenShift Admin Office Hour, please contact us via Discord, Twitter, or come join us live, Wednesdays at 11am EDT / 1500 UTC, on YouTube and Twitch.

Episode 25 recorded stream:

Supporting links for today’s topic:

As we promised on the stream, here is the full list of topics / links discussed related to day 2 node and cluster configuration. You can find last week’s links in the blog post here. Please remember this is only a subset of the information found in the documentation under the “post-installation configuration” section.

Network

Configure secondary+ interfaces that weren’t configured during deployment, preferably using the NMstate Operator
If needed, add DNS forwarders using the DNS Operator
If you’re using OVN Kubernetes and want to use IPSec encryption, enable it
Replace the certificates for API and Ingress with your own
Configure Ingress sharding and scale up if needed
Review optimizing networking and routing docs sections

Storage

If you’re using OpenShift Container Storage (OCS), and/or other hosted storage offerings from the partner ecosystem, deploy them now
Configure default storage class - OpenShift will deploy and configure a default storage class with IPI and UPI. You can customize that storage class based on the infrastructure type the cluster is deployed to.
Configure additional CSI provisioners, for example from partners, as needed for your environment

Monitoring

Utilize the OpenShift monitoring dashboards to make adjustments and decisions as applications are deployed and utilization goes up

Users

Configure authentication and authorization for your cluster
Configure the default Project template, used whenever new Projects are created
Using the Image configuration, add allowed, forbidden, and insecure registries which may be needed for your applications
Prepare to help educate and onboard users by explaining to them how resource allocations work, along with how Pod scheduling hints and Pod priority affect availability

Other links and materials referenced during the stream:

Use this link to jump directly to where we start talking about today’s topic

Questions answered during the stream:

What options do I have if I don’t want to use the built-in load balancer deployed with on-prem IPI?
Can the Ingress controllers be scaled across multiple infra nodes?
When deploying using IPI, how do I use a DHCP reservation for control plane nodes when I don’t know the MAC addresses they’re using before deployment? Red Hat encourages using DHCP reservations for control plane nodes to prevent the IP from changing, which could lead to etcd failing. The answer is both simple and a bit frustrating: convert the lease to a reservation after deployment using your DHCP server’s tools.

If you’re using bare metal IPI and the DHCP lease time is set to “infinite”, then the node will automatically configure a static IP address. But, this only works with bare metal IPI!