This is the slide show Thomas used for his presentation as Open Source Bridge. Open Source Bridge is a new conference for developers working with open source technologies and for people interested in learning the open source way. His presentation explores the fundamental technologies being used by hosting providers, and bridges these concepts with open source development and application deployment. Providing examples of where failure can occur, he also outlined what questions to ask the provider to ensure optimal uptime for applications. Finally the presentation goes over why redundant power and back up power are so critical, and why you should always back up your information. We got a lot of great feedback from the developers who attended the session and decided we should share it with all of you as well. Enjoy!

If you enjoyed this post, make sure you subscribe to my RSS feed!

Portlanders and those of you nearby, I’d like to introduce a conference coming up called Open Source Bridge.

If you are a Network Redux customer we can issue a $100 discount of the $250 registration. As one of the three sponsoring groups with commiter status we also have a couple of free passes reserved for the first two customers of ours to ask.

I will be speaking at a business focused seminar on the 17th called Bridging the Developer and the Datacenter. Huge bonus points for the Network Redux customers who attend the session!

There is quite a large array of fantastic sessions to attend. The first two customers to inquire will get the free passes!

Cheers

Thomas

If you enjoyed this post, make sure you subscribe to my RSS feed!

Ten key items which separate us from the competition…

1. We monitor the health of your virtual private servers on 5 minute intervals. The slightest imperfections of service such as reaching the memory limits you have paid for, or a sudden spike in traffic will flag a system engineer’s attention. We will temporarily raise your resource limits to keep your node running, so you have time to either make the necessary changes to your application or purchase additional resources.

2. We’ll build and deploy any custom configuration you can throw at us. From basic LAMP stacks to MySQL clusters with multiple nodes, we’ll build it for you at absolutely no additional charge.

3. We can scale your 256MB virtual private server to 1024MB or 32GB in real time, without rebooting your node. We can scale your processing power to 8 cores in real time, without rebooting your node.

4. We take nightly system snapshots of your entire server and retain this data for 30 days. You can request a restore of this data at no additional charge.

5. You can rebuild your virtual private server with one of our many templates using our management panel.

6. You can access out of band management to work on your node if it fails to startup or you accidentally lock yourself out while configuring iptables.

7. We can build high availability failover for your apps for as low as $38/Month (2 256MB servers) between our multiple datacenters.

8. Your data is stored on enterprise grade storage architecture, SAS-15K disks in RAID-10 arrays.

9. Every single server chassis in our network contains two power supplies, which are fed from independent power sources, providing true power redundancy.

10. If you have an open source project or app that needs a home, we’ll provide hosting for you free of charge.

We are truly unique when it comes to managed services at an affordable price.  Check out what some of our customers have to say:

https://forums.networkredux.com/index.php?board=4.0

Cheers,

Thomas Brenneke
Network Redux, LLC

If you enjoyed this post, make sure you subscribe to my RSS feed!

Our Portland clients and colleagues may already be aware, but a local establishment (Cube Space) which provides excellent services to technologists in the Portland community is in need of help.

We have pledged and donated $500 to help them reach their goal. More information on this situation can be found here:

http://savecubespacepdx.com

This type of grass roots effort to help out fellow entrepreneurs in the Portland area is one of the reasons we are proud to own and operate a business in this fine city.

Cheers, and good luck to our colleagues at Cube Space.

If you enjoyed this post, make sure you subscribe to my RSS feed!

A company I have admired since their initial inception is Softlayer Technologies, Inc. of Texas.  They provide Hardware as a Service to thousands of customers across the globe.  Unique to their offering is a feature rich portal and API.

Early in their growth we managed servers for approximately ~20 customers at Softlayer.  As our portland based operations grew, and our virtualization offerings matured, this number has since dwindled to three.

Recently, each time we’ve migrated a customer from one of their unmanaged dedicated servers, to our managed virtual private server environment, a member of their rigorous sales team has tried to reach us via telephone and email (usually within 15 minutes of the cancellation request).  Though respectful on the surface, the underlying tone of their sales pitch could easily be construed as “ours is bigger than yours, therefore let the big boys handle the architecture,” as seen in a recent blog post by one of their team members.

Does this type of size still matter?  It would be interesting to see actual metrics on the 20,000+ servers they have provisioned for customers.  CPU usage, physical memory consumption, disk IO and disk space usage.  Too often we’ve brought a customer to our services from a dedicated server to a managed virtual private server, saving them several hundred, in certain cases several thousand dollars per month.  How?  We built a solution that was appropriate to their actual requirements.  We took into account CPU utilization, memory consumption, disk IO, disk space usage, and SLA requirements — using our virtualization technologies and our own network of servers, built a solution appropriate to the customers needs.

I’m losing a tremendous amount of respect for organizations such as Softlayer which seem to believe that strength is in the raw number of servers they can rack.  We build at minimum 8 core servers with 32GB of memory, and we build these on an as need, utility basis.  We’re using our space, power and cooling efficiently and in what the competition is taking credit for as a green manner.

While hardware as a service has its place, more focus should be spent on the quality of service, rather than the quantity of service.  I can say this with due respect, as one of our largest customers was previously a several hundred server customer at Softlayer.  This customer’s new found savings have allowed them to focus their energy less on dozens of dedicated servers (which were unfortunately prone to disk drive failures), and more towards their core competencies.

First impressions are everything.  45,000 server capacity is a marketing point, gorgeous PDFs which highlight this capacity are business intelligence at its finest…

Selling the customer a solution that fits their actual needs and doesn’t constrain them to a physical device that requires downtime to upgrade (or downgrade) is our method of delivering a sustainable platform.

If you enjoyed this post, make sure you subscribe to my RSS feed!