HNHacker News
TopNewBestAskShowJobs

mugsie

1,027 karma · joined May 25, 2012

Graham Hayes

Software / Infrastructure Architect @ Udemy

Previously: Principal Engineer @ Microsoft Azure, Distinguished Member of Technical Staff @ Verizon, Master Engineer @ SUSE Senior Engineer @ Hewlett Packard Enterprise

ex - PTL for OpenStack Designate (DNS Project) (https://docs.openstack.org/designate/latest/)

ex - Member of the OpenStack Technical Committee

[ my public key: https://keybase.io/grahamhayes; my proof: https://keybase.io/grahamhayes/sigs/-W5bXnllXeYREBziC7KFYkLfh0caT8KLU6d7mQ20Tdw ]

submissionscomments
mugsie··on SUSE to Acquire Rancher Labs
Some of that is easily done with Helm and kubeadm, but not all of it, and definitely doesn't scale as you grow the number and size of the clusters running.
mugsie··on SUSE to Acquire Rancher Labs
One thing I would say about SUSE - they are never mediocre. (nor what I would usually consider a behemoth)

The engineering team inside SUSE are exceptional - they do amazing things, and build really interesting features. The product planning / joined up thinking / visionary direction is where they fall down. As a sibling pointed out they have worked on some really interesting stuff, but (when I was there in any case) failed to pull it together into something that could have been outstanding.

mugsie··on SUSE to Acquire Rancher Labs
Red Hat is great - they have done some great things, however -

Heart, Mind and Soul of the Open Source community is possibly a bit of hyperbole.

and - SUSE is (slightly) older than RedHat, so I am not sure you can say RedHat inspired them :)

IBM was inspired by RedHat's earnings (pre acquisition, open source in IBM was .... interesting ... ) and their ability to have a relevant product in the cloud space.

mugsie··on SUSE to Acquire Rancher Labs
I was talking about the GP comparing their OpenStack install experience to something like Rancher, which is not an apples to apples comparison.

- on a side note - OpenStack and Kubernetes are not competitors, they are quite complementary collections of applications, that both have their place in a modern open source infrastructure.

mugsie··on SUSE to Acquire Rancher Labs
Hopefully, but Cluster API relies on something like Rancher (or AKS/EkS/GKE) to do the deployment underneath it - it still kind of outsources the life cycle.
mugsie··on SUSE to Acquire Rancher Labs
K8S is a lot easier than OpenStack to install, but when comparing something like Rancher to OpenStack, it should be compared to something like OpenStack Ansible, or a vendor version of OpenStack (RIP HPE Helion) which were a lot easier than apt-get install openstack is.

K8S has a lot less moving parts - a couple of binaries / containers and etcd. The issue start coming up when you go beyond the single control plane node, and want a HA API.

mugsie··on SUSE to Acquire Rancher Labs
Replacing nodes, helping repairing broken things, upgrading the control plane, upgrading etcd etc.
mugsie··on SUSE to Acquire Rancher Labs
The real value is the Install / Life Cycle orchestration - vanilla K8S has really marked that firmly as "not their problem" - which is the correct thing for them to do.
mugsie··on Apple, Hey, and the Path Forward
OK, that's probably what I remember then
mugsie··on Apple, Hey, and the Path Forward
Can you?

I am pretty sure that the rules about IAP make it very clear you have to charge the same price on and off the app store.

mugsie··on Apple, Hey, and the Path Forward
I don't think you can use any of basecamp features for free. [1]

They make money, that is not up for debate.

(and yes, I know, their platform, their rules, etc etc, but nothing about that says people cannot fight it, and make it politically difficult to let them bleed money from people for a very basic service.)

1 - I have never used basecamp, or looked into it, so YMMV

mugsie··on Apple, Hey, and the Path Forward
> Name one retailer lets you hock your product in their store without paying anything?

Apple. Google. Microsoft. Samsung.

> Who wants an app that you can’t do anything with when you download it?

People who want a client for their email service?

Hell, fastmail is even an email app, that has no in app purchases, and doesn't work without an account. (and is in the app store)

mugsie··on GitHub isn't fun anymore
I have been on GitHub for years, and had no idea there was a trending page at all ... but then again I don't tend to "discover" new repos as a hobby.

(There is a separate rant about using github stars as a metric of, well, anything, but thats a thing for a different day)

mugsie··on Ask HN: Dear open source devs how do you sustain yourself
Day job, at least somewhat related to my open source work. I have been quite lucky that most of my employers have supported my open source work, from time allocated to it during "working hours", to funding conference attendance.
mugsie··on Diagram as Code
> sounds sweet. imagine applying gerrit to infra

Yeah, it is cool - https://opendev.org/ is done that way, as is some of the wikimedia labs infrastructure

I am not certified either - I have just worked in places with ITIL inspired processes, but I can really see a way they can move forward with tools like terraform / pulumi / ansible and git

mugsie··on Diagram as Code
The explicit, declarative model, using tools that don't need a ton of other things deployed.

Ideally the same tool can be run locally and in CI (with the same invocation commands) so that an Op can run tests locally, and even see the potential diff, before putting it up for review.

The advantage of this approach is you can see a diff, and use standard tooling like git / pr reviews to approve changes in day to day, and when the world is on fire, someone can break the glass and run it locally.

Never underestimate the value of peer review on a change to the infra - there is a reason things like ITIL exist, and we should learn from them.

mugsie··on Diagram as Code
I like this a lot - trying to store diagrams that look good, without sticking a visio file into a git repo is something I keep an eye out for.
mugsie··on Diagram as Code
> It’s all HTTP APIs. Build a web UX and store the state in a DB.

Please no. Don't make running a stateful service a requirement for tooling people need to use when the world is on fire. Or needs to load 40Mb of JS to rebuild a deployment, and works only on Chrome 46.012.2039465. Ops/SRE/DevOps/SysAdmin/SysOps/etc are different, and have optimised things to reduce the amount of dependancies.

mugsie··on GitHub's new notifications: a case of regressive design
I think the new one is OK for a quick overview, but I still prefer email. Sieve + this rule[1] files emails from each repo in a folder (following org structure), and I can flag things I need to come back to, mark things I don't care about as read, and have thunderbird pop a notification in the top of my screen for repos I really care about.

(I know I am definitely not the target audience for the page)

1 -

  # GitHub
  if allof(exists "X-github-recipient",
           header :regex "List-Id" "<([a-z_0-9-]+)\.([a-z_0-9-]+)[.@]") {
      set :lower "organisation" "${2}";
      set :lower "repo" "${1}";
      fileinto :create "INBOX.GitHub Code Review.${organisation}.${repo}";
      stop;
  }
mugsie··on A note of reflection after 10 years of “A successful Git branching model”
"Trunk based development"

It is fine, unless you do a lot of back ports or bug fixes for older releases, at which point the version branches are very useful.

mugsie··on Ask HN: Who is hiring? (March 2020)
Microsoft | DevOps Engineers & Software Engineers | Dublin, Ireland - ONSITE | Full Time

Microsoft Azure Global E2E^2 is hiring for two teams in Dublin, Ireland.

Both of the teams: the job says Senior, but we are hiring at all bands.

Team A: The team is responsible for running workloads on Azure and using them to ensure the customer experience is a good one. This role will be building out the infrastructure to collect the telemetry from these applications, and mine the data collected for insights into the Azure platform, and ensuring the whole system stays up and running so that the test workloads can keep pushing data in. There will be an element of on call work - but the on call rota will be spread across 2 time zones, which means reasonable on call hours.

Experience in running a production high volume metrics ingestion system is a plus, along with a steady hand for when things break.

Team 1: This team is responsible for building end to end tested reference architecture for Azure, and is currently working on a reference multi tenant container hosting platform based on Kubernetes (both AKS for in cloud, and upstream K8S for edge / remote sites). The aim is to open source the work.

Upstream open source experience (in any open source IaaS stack, or IaaS management tools like terraform / ansible) is definitely a plus, along with experience in regulated environments (banking, telco or government).

Both teams - apply using this advert: https://careers.microsoft.com/us/en/job/785404/Senior-Softwa...

Any questions - graham(dot)hayes(at)microsoft(dot)com, or via the link above.

mugsie··on Is Arm ready for server dominance?
There is a port of windows already out and running on arm - MS even released a laptop running it ... https://www.microsoft.com/en-us/p/surface-pro-x/8vdnrp2m6hhc...
mugsie··on How to fuck up software releases
I really like how the OpenStack project does releases - it has a ton of tooling, but for people who are deciding what to release, it is super simple.

e.g. - the latest RC of a project I maintain is L10-14 on https://opendev.org/openstack/releases/src/branch/master/del...

by committing that file, the tooling chooses the right commit, does a tag, and builds the python tarballs, and puts them on releases.openstack.org, and for the client libraries, it pushes them to pypi.

it avoids a lot of the footguns of special scripts in repos, while allowing an easy release process.

mugsie··on List of open source applications for macOS
> Great software should be paid for.

Why?

Great software - Linux - is not paid, and it is taking over the hosting world, and the IOT world.

GNU is great software. nano / pico is great software. python is great software. .NET is great software. it is all open source.

mugsie··on List of open source applications for macOS
> because “we want free software” doesn't really translate to pay for hard work and dedication.

That is one hell of a hot take. RedHat, SUSE, Canonical, would all probably disagree. (also, fwiw, so would I - I have been working on Open Source software for 5+ years, and have been paid for it for the entire time)

> as most are either contract workers or just volunteers

The linux kernel called, said something about having a large amount of fully paid, long term staff.

I get it, you don't like free software - but there is plenty of free software that people are paid to maintain and develop, and it has been a great benefit to software development as a result. Even paid for software is probably using some open source components, (runtimes, languages, compilers, libraries to name a few of them).

mugsie··on The secret life of open source developers [video]
Those sort of emails are really hard to deal with, especially when it is something you spent a lot of time working on.

I like the idea from the talk of just blocking them outright.

mugsie··on The secret life of open source developers [video]
> in particular, you have no power to make the users polite.

So very (and unfortunately) true. However they also have no power to make you do the thing they need straight away, or ever.

> Even if you put hard work in it, you sacrifice a bit of your life.

And this is bad. We need to move the needle away from this. It is incredibly exclusionary

I am a younger person, no kids, in a job where I was able to negotiate for open source maintenance time. By making "sacrifice" a requirement it makes it off putting for someone with kids, or working in a job that has them working long hours, or working multiple jobs.

> But the user, well, they just are what they are, software can't fix that.

No, software can't fix it. but highlighting the problem, and letting people possibly see what they are doing is not useful may change the user.

I have been a maintainer of a project for a long time, and I get everything from people pinging me in our IRC channel, to DMing me, to asking if I give them my skype ID so I can call into a bridge to help them debug an issue at midnight. We need to highlight this is not OK.

Personally, I really like the PowerDNS rule for open source support [1] - it sets expectations, and the base requirements from someone looking for help.

> the success, at least to me, was measured by the number of users

That is a valid metric, but just one. I personally measured the success of the project based on what people were using it (e.g. I would rather wikimedia vs 20 startups), but that metric is really person specific.

1 - https://blog.powerdns.com/2016/01/18/open-source-support-out...

mugsie··on Apple’s New Mac Pro to Be Made in Texas
I just spec'd out a HP Z6 to the min spec of the Mac Pro - it came to about $4000 (and that is EU prices, I am sure the US prices are less) - so 20% difference.

An maxed out z8 is $50k - but that is way over what most people need, or you can do with a Mac Pro

If you look at other manafacturers - you can get insane performance for much cheaper.

2 x AMD Epycs (64 cores per CPU, 128 cores / 256 threads in the box) 1TB RAM 2x2TB SSD 2 x Quadro RTX 6000

Is $38,000. It is not even in the same sport, let alone league as the mac pro

mugsie··on Create FreeBSD Current OpenStack Image on OVH Public Cloud
I am biased[1], but the OpenStack story is pretty good. Running it is always going to be more than "apt install openstack" - but there has been pretty good improvements in installers.

Juju (linked below), OpenStack-Ansible, or openstack helm all make deployment possible.

The end user story has been pretty consistent for a while, with additions for newer features, and high performance hardware always in the pipeline.

A lot of the horror stories from the early days were people expecting free, fire and forget ESXi - which it is not.

1 - I have been a project team lead on one of the sub projects for ~ 4 years, and more recently sit on the Technical Committee.

mugsie··on GitHub Outage Currently Ongoing – Issues, PRs, Notifications
Modern apps are designed to deal with part of them being down - it is the core upgrade strategy.

VMS clusters were (and modern equivalents are) awesome, but they have a different speed of innovation / development.

It is a trade off - perfect uptime, but that new feature could take a year, or 99.9% uptime, but it can be written and deployed tomorrow.

← PreviousPage 3 of 12Next →