Our Location – Reno, Nevada

Reno, NV is about a 2 hour drive from Sacramento, CA, 4 hours from San Francisco, CA, and 8 hours from Las Vegas, NV. If you’re flying in, we’re about 8 minutes away from the Reno-Tahoe International Airport. We’re close enough to California to tie into communications infrastructure without being affected by events such as the 2000-2001 California electricity crisis or the 1996 Western North-America summer blackouts, and we’re safely away from high-risk California earthquakes, making Roller Network the perfect location for a disaster recovery or backup site.

For our local customers, our data center is easily accessible from major roads (McCarran, S. Virginia, US395/I-580, Longley/Double R/Airway) within city limits. We’re across the street from the new Northern Nevada Sierra Medical Center.

The History of Roller Network

Roller Network was founded in 2004 by Seth Mattinen, then a Computer Science student at the University of Nevada, Reno, as a side project away from analyzing algorithms. Our first service was a “free for all” backup mail exchanger. As the number of features grew, a donation was required to access new features. The customer base continued to grow through word of mouth to the point where donations simply weren’t keeping up with demand, so the free model was discontinued in December 2007. Today, Roller Network is home to thousands of mail and DNS domains with a focus on leaving as much control as possible in the hands of the customer.

Roller Network was originally created to address the needs of small and medium businesses and technically savvy customers. We are one of few mail service providers to break from the per-domain or per-user pricing scheme and offer single price point mail services. We are also one of the first to offer IPv6 enabled hosted mail services.

In Fall of 2009 we left our longtime home in Sparks, NV and sought to build a facility-centric colocation datacenter in Reno, NV. Since the beginning our mail services were self hosted due to a lack of proper local hosting options, so we elected to create our own facility and eventually opened it up for retail colocation. With our colocation facility, Roller Network was able to offer a range of services that weren’t found in Northern Nevada at the time. Our facility is designed to focus on equipment first and foremost. In 2010, we extended our network through the Truckee Meadows through a handful of microwave sites. From those sites we offered Metro Ethernet and wireless broadband for businesses. In 2020 we started to scale back our basic ISP services to focus on colocation and point-to-point microwave.

Roller Network strives to provide a high level of service in operating our own facility: power distribution, backup generator, site controls, switching, and routing. If something is not quite right, we can fix it. If you have a question, we can answer it. If we need something, we can do it. We diversify our fiber access with multiple physical and logical providers with independent paths. We employ online UPS systems. Our generator is an automatic diesel and gets exercised, on load, twice a month, proving its availability. Our cooling systems are redundant with an environmentally friendly economizer system. If there’s a problem – since no system is perfect – we believe in transparency during the resolution process.

History of IPv6 Support

Roller Network has been IPv6 enabled since August 29, 2005 when we brought up our first IPv6 BGP session with Sprint.

  • 2015-09-24 – ARIN Reaches IPv4 Depletion
  • 2012-09-18 – ARIN Enters Phase Two of their IPv4 Countdown Plan
  • 2012-09-14 – Europe Officially Runs Out of IPv4 Addresses
  • 2011-04-14 – APNIC First to Run Out of IPv4 Addresses
  • 2011-02-03 – Free Pool of IPv4 Address Space Depleted
  • 2011-02-01 – Significant Announcement 3 February – Watch it live!
  • 2010-12-24 – AAAA records for incoming mail have been published in DNS. This marks complete support for IPv6 in our services.
  • 2010-12-21 – We’re putting the finishing touches on IPv6 support for our incoming mail services.
  • 2010-11-30 – We have started migrating IPv6 services away from our test and development network and into their final home within 2607:fe70::/32. During the transition period the existing services will listen on both their old and new addresses.
  • 2010-08-31 – 6to4 Relay Activated
  • 2010-07-02 – We’ve turned off our supplementary tunnels to Hurricane Electric and LavaNet, having been replaced by native connectivity from Global Crossing. We would like to thank them for their service during our issues with Verizon.
  • 2010-06-10 – Our native IPv6 with Global Crossing will be turned up Monday at 12:00 local time.
  • 2010-05-08 – Mail forwarding is now IPv6 enabled. If a forwarding destination is reachable via IPv6, we will attempt to use it.
  • 2010-02-01 – Send Mail Using IPv6!
  • 2009-11-06 – Outbound Mail Now IPv6 Enabled
  • 2009-10-04 – We made Slashdot! Technology: Verizon Refuses To Provide Complete IPv6
  • 2009-10-03 – Verizon Refuses to Provide Complete IPv6
  • 2009-08-11 – A problem with our redundant mail hosting servers was discovered in testing – the IPv6 address won’t fail over quickly due to limitations in Linux libnet. Basically, our routers won’t see the change in the IPv6 physical layer destination. We can execute a “clear ipv6 neighbors” to fix it, but it requires manual intervention and is not something we should have to do.
  • 2009-07-27 – We may have a solution to the “IPv6 addresses in the logs” problem by storing them as DECIMAL(39). We store IP addresses in their integer value (as opposed to a string) to make searching by IP possible.
  • 2009-06-06 – Our tunnel to Hurricane Electric was trapped by an OSPF bug; they fixed it by resetting the tunnel interface.
  • 2009-06-04 – We discovered a routing black hole on IPv6 traffic coming to us through Hurricane Electric. We don’t know how long it’s been happening, but we shut off our BGP connection to them and restored connectivity. (Our inability to reach whois.arin.net via IPv6 tipped us off. Further testing revealed a SixXS traceroute died at HE.)
  • 2009-05-06 – It’s been a while since our last IPv6 update, but we haven’t stopped working on it. We are attempting to acquire a dual-stack connection with Verizon as part of our move to a new facility.
  • 2008-12-11 – Testing for enabling IPv6 for our mail services has begun.
  • 2008-11-21 – We’ve added IPv6 access for our hosted mail boxes under “mailbox.ipv6.rollernet.us” and to webmail at “https://webmail.ipv6.rollernet.us”.
  • 2008-10-21 – A Treo 700wx on Sprint’s network can access this site (and others) using 6to4.
  • 2008-10-15 – So far there don’t appear to be any adverse effects from using A/AAAA records on our ns2 and ns2-auth records.
  • 2008-10-05 – We’re going to try giving “ns2.rollernet.us” and “ns2-auth.rollernet.us” AAAA records. Since DNS lookups aren’t interactive, broken IPv6 resolvers should fall back. (Some root servers have A and AAAA records as well. This is the same thing.)
  • 2008-10-01 – Enabled IPv6 for our Primary DNS service.
  • 2008-09-30 – Enabled IPv6 for our Secondary DNS service.
  • 2008-09-02 – We have completely dual-stacked our network core.
  • 2007-11-19 – First IPv6 network under 2620:0:950::/48.
  • 2005-08-26 – Early testing under 3FFE:2900:1118::/48 through Sprint.