#kubernetes (2019-11)

kubernetes

Archive: https://archive.sweetops.com/kubernetes/

2019-11-22

2019-11-21

Thomas Burton avatar
Thomas Burton

What GitOps tools do people use here. I have been looking into FluxCD and ArgoCD . Interested to hear about other peoples experiences with any other relevant tools. https://github.com/fluxcd/flux https://argoproj.github.io/

fluxcd/flux

The GitOps Kubernetes operator. Contribute to fluxcd/flux development by creating an account on GitHub.

Get stuff done with Kubernetes | Argo

Kubernetes native workflows, deployments, CI, events

roth.andy avatar
roth.andy

Harness looks promising. My team is working on testing with it. Will report back later.

fluxcd/flux

The GitOps Kubernetes operator. Contribute to fluxcd/flux development by creating an account on GitHub.

Get stuff done with Kubernetes | Argo

Kubernetes native workflows, deployments, CI, events

roth.andy avatar
roth.andy

Also CodeFresh looks nice for setting up simple stuff since you can trigger off of pushes to docker registries (along with a bunch of other stuff)

Thomas Burton avatar
Thomas Burton

Thanks Andrew. Let me know how you get on

Chris Fowles avatar
Chris Fowles

We’re currently using flux in a new eks platform we’re building out - just starting to enter developer testing now and things are looking good

pianoriko2 avatar
pianoriko2

Hi, has anyone used k8s and route53 on gov cloud?

roth.andy avatar
roth.andy

LevelUp has some open source material that they have published. https://dccscr.dsop.io/levelup-automation/aws-infrastructure

2019-11-20

github140 avatar
github140

Hi, does anybody know if images of the OpenShift internal registry could be pulled cluster internally and pushed to an external (like Nexus) repository?

2019-11-19

curious deviant avatar
curious deviant

Hello,

What are folks using as their helm chart repository ? I began by using an S3 bucket in our shared AWS account. But would like to hear if there’s something better out there in terms of user/developer experience

2019-11-15

roth.andy avatar
roth.andy

Somebody else go through this list and tell me if any of them sound fishy. I want to see if your list matches my list.

https://learnk8s.io/production-best-practices/

Kubernetes production best practices

This document highlights and consolidates best practices for building, deploying and scaling apps on Kubernetes in production.

:--1:3

2019-11-14

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)
roth.andy avatar
roth.andy

Did you see he is doing an AMA? I’ll find the info

PePe avatar

there is a typo there….I can’t unseen it

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)

This is kind’a rad coming from the DoD.

loren avatar
loren

“fast” in the DoD meaning less than 5 years?

roth.andy avatar
roth.andy

nope, like ~180 days

roth.andy avatar
roth.andy

There’s already an $800 million contract called Cloud One to migrate a ton of apps

loren avatar
loren

we’ll see

2019-11-13

Taras avatar
Taras

Hi guys, Trying to parse kubernetes logs with fluent-bit helm chart form here: https://github.com/helm/charts/tree/master/stable/fluent-bit Stuck in configuring outputs. What I need is to have a few outputs sending logs to different indices into AWS ES based on kubeernetes namespace. If anyone happened to have worked with this helm or similar issue help is welcomed.

helm/charts

Curated applications for Kubernetes. Contribute to helm/charts development by creating an account on GitHub.

2019-11-12

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)
quay/quay

Build, Store, and Distribute your Applications and Containers - quay/quay

2019-11-11

Nelson Jeppesen avatar
Nelson Jeppesen

I’ve been unable to find info how to switch a kops 1.13 cluster from single to multi-master; All the documentation I’ve found provides steps before the switch to etcd-manager. I’m having issues even connecting to etcd with etcdctl

Can anyone point me in the right direction?

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)
aws/aws-node-termination-handler

A Kubernetes DaemonSet to gracefully handle EC2 Spot Instance interruptions. - aws/aws-node-termination-handler

Pablo Costa avatar
Pablo Costa

I think that for while it is better to stick with https://github.com/kube-aws/kube-spot-termination-notice-handler. It lacks the features of asg detach and notifications. ASG detach improves a lot the recovering time making the interruption almost seamless

kube-aws/kube-spot-termination-notice-handler

A Kubernetes DaemonSet to gracefully delete pods 2 minutes before an EC2 Spot Instance gets terminated - kube-aws/kube-spot-termination-notice-handler

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)

@Igor Rodionov

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)


If a termination notice is received for an instance that’s running on the cluster, the termination handler begins a multi-step cordon and drain process for the node.

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)

it’s not spot specific

TBeijen avatar
TBeijen

Asg rebalancing seems a use-case too. (If not disabled or having an asg per az)

:--1:1

2019-11-08

johncblandii avatar
johncblandii

anyone playing w/ https://skaffold.dev/?

Skaffold attachment image

Easy and Repeatable Kubernetes Development

pianoriko2 avatar
pianoriko2

Does any one know how to use kops on gov cloud? I can’t get the DNS right. Any help is appreciated

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)

I thought it wasn’t yet supported

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)


While there are DNS servers in the VPC, there is no Route53 service (API). This breaks many devops tools that make the assumption Route53 is / will be available (kops w/ kubernetes, for example). To be fair, some tools (like kops) provide an alternative to Route53 for bootstrapping the cluster, though our testing found the features to be buggy and not yet production quality. We worked around the need for Route53 by deploying our own self-healing and automated DNS solutions. A future post will dive into the details of our Route53 replacements.

roth.andy avatar
roth.andy

The article is old, though some of it still holds up. Route53 in particular is now available in GovCloud

roth.andy avatar
roth.andy

*now. It is available now

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)
Intro to Devops on GovCloud attachment image

If you have strict compliance criteria that require you to use AWS GovCloud, there are some obstacles you will encounter that we will help you address.

pianoriko2 avatar
pianoriko2

Someone told me of a work around with freeipa with your experience do you think it is a good option?

roth.andy avatar
roth.andy

It works, one of my colleagues is doing it. They said it was janky though. My company is going to be doing a TON of work in the very near future with K8s in GovCloud so I’m looking for more information on the subject as well

2019-11-07

2019-11-06

Alex Siegman avatar
Alex Siegman

Just saw this today, thought it was neat: https://github.com/linki/chaoskube it’s been around a while, guess i’m late to the party lol

linki/chaoskube

chaoskube periodically kills random pods in your Kubernetes cluster. - linki/chaoskube

roth.andy avatar
roth.andy

Cool!

linki/chaoskube

chaoskube periodically kills random pods in your Kubernetes cluster. - linki/chaoskube

Jord avatar

If anyone is interested to test out Magic Sandbox, DM me and I can get you premium access for a week to explore! > https://msb.com

:--1:1
Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)

@Pierre Humberdroz shared in #office-hours today an AWESOME new tool he’s working on to diff helm releases

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)
Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)

UI still rough, but value immense

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)

you can compare the changes between any 2 releases

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)

(see url syntax)

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)

@Jeremy Grodberg @Igor Rodionov @Jeremy Grodberg

Pierre Humberdroz avatar
Pierre Humberdroz

Thanks for sharing @Erik Osterman (Cloud Posse),

The idea was to validate this today with a little working prototype if you have feature ideas let me know. Currently the main benefit to hub.helm.sh is that you are able to compare two chart versions. Other Features that I have planned are:

  • Notification on releases
  • Notification if a new helm chart is added with a keyword you are looking for.
cool-doge1

2019-11-05

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)
Kontena Lens - The Smart Dashboard for Kubernetes attachment image

The only management system you’ll ever need to take control of your Kubernetes clusters.

Joe Niland avatar
Joe Niland

I’ve been using it with Minikube. It’s a much nicer/faster experience than the dashboard addon.

Kontena Lens - The Smart Dashboard for Kubernetes attachment image

The only management system you’ll ever need to take control of your Kubernetes clusters.

Joe Niland avatar
Joe Niland

The only thing I’ve found that’s lacking is the UI controls around logs

1
Chris Fowles avatar
Chris Fowles

been using it for the past few days with a dev eks cluster and have found it very nice so far

johncblandii avatar
johncblandii

Toying w/ it, but it doesn’t seem like it supports iam auth. Going to tinker w/ it on minikube.

Chris Fowles avatar
Chris Fowles

I’ve got it working now with IAM Auth - what issues are you having?

johncblandii avatar
johncblandii

I didn’t really try very hard. I just selected my cluster map and it failed so i gave up. lol. I was just tinkering anyway.

1
johncblandii avatar
johncblandii

Ok…it works fine, @Chris Fowles. It helps if you use the right config AND your AWS profile actually has access.

Chris Fowles avatar
Chris Fowles

lol

1
Chris Fowles avatar
Chris Fowles

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)
1
Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)

Looks like it’s now dead.

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)

Just when I started playing with it.

Chris Fowles avatar
Chris Fowles

you killed it!

Chris Fowles avatar
Chris Fowles

johncblandii avatar
johncblandii

Well that sucks. Just saw this. lol. I use this every day.

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)

Ya, it’s a bummer. But they are working to open source it.

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)

johncblandii avatar
johncblandii

nice.

Cameron Boulton avatar
Cameron Boulton

Have you played with it at all @Erik Osterman (Cloud Posse)?

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)

nope, first i heard of it was today

Erik Osterman (Cloud Posse) avatar
Erik Osterman (Cloud Posse)

via @Daren

2019-11-04

s2504s avatar
s2504s

Hello ALL I’ve found this feature https://github.com/zalando-incubator/stackset-controller Does somebody use it? What is your feedback?

zalando-incubator/stackset-controller

Opinionated StackSet resource for managing application life cycle and traffic switching in Kubernetes - zalando-incubator/stackset-controller

s2504s avatar
s2504s

Is it production ready?

2019-11-01

Jeremy Grodberg avatar
Jeremy Grodberg

An issue has been opened to track the fix for the CFS scheduler bug in CoreOS. People using CoreOS to host Kubernetes may want to track this: https://github.com/coreos/bugs/issues/2623

CFS scheduler bug throttles highly threaded I/O blocked applications in Kubernetes · Issue #2623 · coreos/bugs

Issue Report Bug Container Linux Version NAME="Container Linux by CoreOS" ID=coreos VERSION=2191.5.0 VERSION_ID=2191.5.0 BUILD_ID=2019-09-04-0357 PRETTY_NAME="Container Linux by Core…

:--1:1
    keyboard_arrow_up