IIRC there are already 70(?) million WP sites, with 45(?) million not hosted on wordpress.com. I wouldn't be surprised if this built in git deployment makes that number grow much more quickly. It sounds like a lot of people already treat WP like a tech stack. So, this could basically be treated like heroku.
Yup, "treating it like Heroku" is exactly our idea! They showed the way and we're happy to follow their lead and bring that functionality to the WordPress community.
I thought about moving but I couldn't get past them using 'visitors' as a tier function in their pricing and also claiming 'unlimited' bandwidth those things are opposites to me. Is that unique visitors or page views?
That's a really great question, and a lot of people have asked it. So we had Jason write a blog post to explain how we count visits.
TL;DR "We take the number of unique IP addresses seen in a 24-hour period as the number of “visits” to the site during that period. The number of “visits” in a given month is the sum of those daily visits during that month."
Love the announcement. But a followup question: if WP Engine has excellent built-in caching and optional CDN, why meter so strictly by visits?
It reminds me of A/B services that scale similarly: by the time you get up to 100k or 250k visits per month, you also get 50 domains and 1000 simultaneous tests and all sorts of unnecessary stuff.
I'll be interested in seeing how this works. I only occasionally build sites on WordPress, but every time I do, I struggle to come up with a reasonable deployment process.
Being able to update plugins and the WP core through the admin panel makes it easy for your server and source code to become out of sync. The configuration file also needs to change for production databases. From what I've seen, there's no agreed upon way to deal with these issues.
Taking a look back on all technologies and related configurations I touched (Apache, MySQL, Wordpress, Git), I'd agree it was worth the time. I'm pretty familiar with the basic Git functions (typical pull, push, merge, rebase, etc), though it was my first hands on experience with Git hooks. Git hooks rock.
Agreed. This makes a big difference in how people can deploy WordPress now. Next time you're doing research, you may want to take a look at nginx and how it works with Apache / what the tradeoffs are. It's a blazing fast server technology.
We switched our Entrepreneurs Unpluggd hosting to WP Engine 4 months ago and haven't had a single issue since. Our previous host was atrocious with insecure support, consistent downtime, and an overall unprofessional demeanor.
This sort of innovation is what I love to see with a web host. Keep it up Jason and team!
Disclosure: WP Engine is the hosting sponsor for Entrepreneurs Unpluggd.
And, ironically, their site is down. That's a pretty bad sign for a service that's supposed to provide a more performant installation of Wordpress. I hope it's actually something on the hosting side of things (or maybe they don't use WP for their main site...)
It wasn't working for me before, but appears up now that I'm at home. I was on Georgia Tech campus (which I believe peers with Level3 for most commercial traffic), if that helps. Might have been a backend/infrastructure thing (like I hoped).
Yeah, our DNS provider Zerigo had a nasty DDOS attack over Sunday that many sites got caught up in, including us.
As you can imagine, mitigating DDOS on a nameserver is tricky due to the nature of the requests coming in - harder to Deep Packet Inspect and scrub, or easily identify patterns of bad traffic access to zero-route.
Yup, we're switching to multiple providers. The fact that there were already multiple geographically redundant IPs should have averted most issues, but someone made a highly-coordinated attack on the provider at all their IPs.
We learn and move on. I wonder how many other startups are going as far as using DNS from multiple providers? I still see many startups using Go Daddy for their DNS... :/
Or perhaps c) there's one of many reasons why the site wasn't reachable at the time from the university campus where the person reported in from.
They went home, it works from their home ISP. We're not seeing any other reports of downtime or unavailability across any of our monitoring. No one else here is reporting "down for me too".
The fact that our site was unavilable for them on their university campus, while still concerning, hardly equates to "we don't know much about WordPress" or "our service is bad".
When you get hit with a heavy load and go down, it is an huge issue for a web business. The whole point is to drive large amounts of traffic, and so if a marketing campaign is successful you're paying for nothing.
Apparently you don't think that web based businesses having downtime is an issue. Worse yet, your casual and downright confrontational attitude when confronted with actual evidence.
Put yourself in their shoes for a minute, and try building a company that is pushing tech boundaries.
Not saying you aren't, but based on your tone it would indicate that a) You don't know what it's like to roll out something that generates a lot of interest in a short-period of time, or b) Have forgotten what it's like to be in that situation.
No matter how prepared you are, dealing with a huge burst of traffic in a very short period of time is the quintessential "hard problem" for web developers - especially in modern web businesses where anything can cause the site to slow (e.g. an expired cache on a node on a CDN that happens to be slow/down during the window you launched, that also happens to be serving some asset other aspects of your site rely on - css/js/etc.).
Even Google, Amazon & Apple, who have boatloads of cash and hundreds of man-decades of experience building web infrastructure still have issues like this on a launch day.
So cut them some slack.
When it's your turn to be in that position, we will do our best to return the favor.
Since it isn't clear from the article: Is this "git deployment" just for the code components of WordPress (e.g, core upgrades, themes, and plugins), or does it cover content in the database as well? The latter would be much more interesting and useful.
We're recommending that core not be part of the deployment, since that could conflict with normal versioning of WordPress. It covers files and database. Check git.wpengine.com for the super in-depth coverage.
I have been deploying a Wordpress site using capistrano and git for a while now. Seems to work pretty well and it is configurable if you like that kind of thing. This approach is a little more hands on than wpengine, but it provides reliable site deployment.
We also recently released a Github project for doing this called "WP Stack". Git, Capistrano, CDN rewrites, and staging environment all in one toolkit. knewt.ly/N119BM
That's definitely why you want someone like WPEngine doing this for you; Jason and the team there is fantastic and they rival Automattic for their WP knowledge. Out of the box on WPEngine is astronomically better than out of the box on a regular hosting company.
I don't understand the use case for this at all. I think they're trying to emulate Heroku's git-based deployment, but do WP users even use git? Is Wordpress even designed for continuous deployment?
Also, their git site has a bit: http://git.wpengine.com/