gbrayut

This is a great write up from earlier this year, and from the screenshots it looks like it was before our new site design launched. That was a pretty massive modernization effort that included some site performance improvements. We also now have a sizeable team (100+) focused on some of the other details from the article. Brotli for instance went into A/B testing this month.

There are a lot of moving parts and changing priorities at a company the size of Walmart, and development and deployments can be slower than at smaller shops. But there definitely is active progress on improving our site speed.

And if you like working on any of these kinds of systems we are always hiring!

A previous employer used their identity server for a SSO system. It did what we needed it to and seemed like it was well written.

That seems like a strong statement that I didn't see any backing for in the blog post. Care to expand? I haven't seen much settling in the k8s space. In fact it looks like things are still heating up with many companies pivoting in that direction.

https://github.com/StackExchange/dnscontrol

https://www.usenix.org/conference/srecon17americas/program/p...

The "Presidential" alert is the level of alert. It's been named this for over a decade, but none have ever been sent.

The levels are :

AMBER alert

Severe alerts

Extreme alerts

Presidential alerts

If you do whois lookups against the top 50 websites you'll see a lot of them use a small set of registrar's. But not all of them accept small businesses.

Keep up the great work!

I gave a quick intro to nsq talk at our local go meetup recently. If you haven't used NSQ before I highly recommend giving it a try.

https://docs.google.com/presentation/d/1e9yIm-0aNba_H1gX_u7D...

To be clear, the HAProxy+vrrp+dns is often a better solution, but this describes an interesting design for a load balancing system that can handle many orders of magnitude more traffic and have maintenance without breaking established connections (one of it's core design features)

"if I have learned anything in my career, it is the shocking effectiveness of building ... literally the stupidest thing that could work. (And then iterating on it for a decade.)"

https://twitter.com/codinghorror/status/1026332543153389569

Keep up the great work!

It works well for a large company with many moving parts and a need for Enterprise wide policies/quotas/etc, but has some rough edges and leaky abstractions. Kubernetes is a more mature and widely used system that has many of the same features, just based around containers instead of VMs.

Disclosure: I work for Walmart Labs on our CDN / Edge Platform team.

https://www.roccat.org/en-US/Products/Gaming-Mice/Tyon/

We also went on our honeymoon (Turks and Caicos, highly recommended) and vacations in Portland, Seattle, San Francisco, and Florida.

Shows how many levels of scaling are required to serve traffic at large scale.

There can be a lot of edge cases, and inevitability things will change in the future. Centralising the work of parsing /proc files goes a long way and helps keep things sane for maintenance.

Of course that doesn't stop it from happening, but at least now we can send notice and take action when we find scrapers.