Accept

We use cookies in order to save your preferences so we can provide a feature-rich, personalized website experience. We also use functionality from third-party vendors who may add additional cookies of their own (e.g. Analytics, Maps, Chat, etc). Read more about cookies in our Privacy Policy and Terms of Service. If you do not accept our use of Cookies, please do not use the website.

Gutenblog: redundancy

<  Blog Home
April 2017
April 2017
By

We've completed phase 2 of our upgrades which gives us new servers that are faster, more secure, have more redundancy and are running newer operating systems for better long term support.

Faster with HTTP 2.0

This upgrade also enables HTTP 2.0 for all websites with SSL (Secure Certificates). HTTP 2.0 allows pages to load much faster by enabling multiple concurrent connections at once. That means if your page has 50 resources that have to be loaded (e.g. HTML, CSS, Javascript, and image files), instead of loading one at a time, all 50 can load asynchronously.

Our Apologies for the Downtime

We did experience two problems during the upgrade that affected some visitors to a few sites:

1. Intermittent Loading Problems: Our security software on the new web servers mistakenly detected that our Load Balancers were performing a DDOS attack on them. As a result, our web servers intermittently throttled the connections, which caused some resources to fail to load (e.g. the CSS or an image ...


March 2017
March 2017
By

Our planned web server infrastructure upgrades were completed successfully. It required some key changes to the Gutensite platform to allow sessions to persist across multiple servers, and for custom files to be accessible on all servers. Big thanks to Matt for taking on this task.

The new infrastructure allows us to do true "horizontal scaling", with load balancing and much better redundancy. That means that your website is hosted on multiple servers at once. Every page load is distributed across a network of servers that are balanced based on load. If one web server becomes unresponsive, or needs to be taken offline for maintenance, the other servers will take up the load and there should be no interruption or slowness. It's an exciting upgrade for us.

What's Next?

Our next immediate infrastructure plans are to migrate the Gutensite web and database servers from our current datacenter at Rackspace to our new datacenter at Linode. Rackspace is a widely respected hosting service, ...


August 2013
August 2013
By

As part of our long term server infrastructure upgrades, we are integrating with Rackspace's Cloud Files to store all your images, documents and multimedia. Cloud Files provides 3 redundant copies of your files (in case any single version gets corrupted or is on hardware that fails), and it delivers those files to your visitors through a CDN (Content Delivery Nework) that caches a copy of the files closer to the visitor which speeds up your site load times. It also allows us to offer unlimited amounts of scalable storage and fast bandwidth at competitive rates.

We're happy to announce that last week we successfully integrated all multimedia (video and audio files) to Cloud Files. This week we will move over documents (e.g. pdfs and other files from your downloadable resource library), and then we'll move over all your images.

Once we complete the Cloud Files migration (hopefully this week) we will begin the web server migration to Rackspace's Cloud Servers. This will provide much ...


July 2010
July 2010

You may have noticed that your site was unavailable last friday (July 16th at around 12:30 PM PST). The downtime lasted about 30 minutes and was caused by a hardware failure with one of the core routers at the data center that hosts our servers. This affected their entire data center, including our server. It was not a problem with our server per se, but a network problem that slowed or stopped traffic from getting to our server. We take uptime very seriously, but these things happen, unfortunately, despite redundancies and backups in the network.

Our secondary backup server (located in California), detected that our main server (located in Montana) was unreachable and promptly redirected all traffic to itself. However, at the moment the backup server is not a full sync of our main server, so all it can display is a nice error message that explains the problem (not the entire website, but better than a white screen). In the future, though, we plan on creating a fully redundant ...