“Tiers” in the telecom world tends to have multiple definitions, depending on who you ask. What tier a bandwidth carrier is, is loosely defined at best and everyone has their own opinion of what tier a carrier might be. Here is my opinion:
Tier-1: [may be considered an RBOC or LEC – Regional Bell Operating Company or Local Exchange Carrier]
Tier-1 is a network in which only settlement free peers and customers are serviced. The network operator pays for none of it’s transit.
Tier-1 is the optimum network backbone for medium to large businesses with critical reliability, stability, and scalability requirements.
Tier-1 can be an advantage when it comes to handling DDoS attacks: if you ask/configure your Tier-1 provider to null-route an IP they will implement the null-route at their borders, so there is no point of saturation.
Examples of US Tier 1 carriers:
AT&T (formerly SBC, Bell South, Southwestern Bell, Ameritech)
Verizon (formerly MCI and UUNET)
Level 3 (recently merged with Broadwing)
Tier-2: [may be considered a CLEC (Competitive Local Exchange Carrier); has their own network, but also resells tier 1]
Tier-2 is where the network operator buys all or some of it’s transit from a Tier-1 and resells it.
Tier-2 can be an advantage if you need someone to provide quality bandwidth, and especially if your need is a single install location. If you buy from Tier-1 #1 in New York, and #1 has a problem with its Tier-1 #2 peering router in New York, then all your traffic from you to #2 may be affected. Your ability to shout at #1 and get them to fix it will be limited, especially if the problem is with #2’s border router. A good Tier-2 will monitor its upstreams and their peering points for trouble, and take measures to ensure that it doesn’t affect their customers. Even if you need multi-location installs, buying from a good Tier-2 can be useful.
A Tier-2 that only responds to severe problems (e.g. total outage of an upstream link) is no more useful than a Tier-1 to someone who has multi-locations.
Pricing from Tier-2 ISPs is often cheaper at the low-end (e.g. T1). Tier-2’s will often beat the tier-1’s in pricing “access services”. But if you buy in the hundreds of megabits, a Tier-2 is likely to quote much higher than a Tier-1.
Tier-2’s are usually smaller companies, and are better able to “make deals”, or recognize bundling of contracts, write custom SLAs (Service Level Agreements), trench fiber to your location in exchange for that signed contract, etc. Unless you buy multiple gigabits from your upstreams, if you want to bundle contracts with Tier-1s, you will probably end-up doing it through a wholesaler or other buying mechanism.
Examples of US Tier-2 carriers:
Time Warner Telecom (recently merged with Xpedius)
Eschelon Telecom Inc
AOL Transit Data Network
Tier 3: [wholesalers /resellers of tier 1 and 2 networks]
Tier-3 are downstream customers of Tier-2’s. Tier-3 may give you what looks like a good price…..but longterm reliability, performance, and scalability will likely suffer. For piece of mind for your business a Tier-1 or Tier-2 are better choices is almost every case. However, PowerNet Global is a big exception to this rule due to their solid infrastructure and relationships with major players.
Examples of US Tier-3 carriers:
Access One Inc.
Whatever business application you need met be sure to consider what tier your available bandwidth providers are in your purchasing decision. Ignoring this factor in your deliberations may result in less than optimum implementation and satisfaction in the end. Make a smart business decision…..leave nothing to chance.