Frequently Asked Questions

MOREnet's Backbone

How is MOREnet's network arranged?

MOREnet has five hub locations: Columbia, Kansas City, Springfield, St. Louis and Stoutland. MOREnet has a 20-year Indefeasible Right of Use (IRU) on two strands of fiber optical cable that connects the five hubs together in a ring plus two additional strands that bisect the ring by connecting Columbia to Stoutland. MOREnet uses Fujitsu optical Equipment to send signals over the fiber. The MOREnet backbone segments are configured as 10-gigabit per second Ethernet circuits. The ring plus bisector topology allows MOREnet to provide redundant paths for each segment giving the backbone excellent protection from fiber cuts or equipment failures.

What equipment do you use at the core?

MOREnet has Cisco 10000-series routers and/or Cisco ASR 1000-series routers at each hub to aggregate customer circuits and a Cisco 12000-series router to provide backbone routing. In St. Louis and Kansas City, MOREnet also has Cisco ASR 9000 routers to connect to the Internet. Each hub also has a DNS server.

What equipment do you use at the edge?

At member sites, MOREnet uses a variety of edge routers depending on the type and speed of the member’s connection. Currently MOREnet is using the following for edge equipment: Cisco 1841, Cisco 2811, Cisco 2821, Cisco 3845, Cisco 7304 and Cisco ME3400. Some older models of Cisco routers are still in use but are in the process of being replaced.

External Connectivity

What providers does MOREnet connect to?

MOREnet currently has two Internet providers: Telia Sonera and Level 3. MOREnet connects to Level3 in St. Louis and to Telia Sonera in Kansas City. Both connections have a capacity of 10 Gigabits per second.

What is Internet2?

Internet2 is a high-speed research network. It was originally intended to connect together colleges, universities, and government institutions whose research would benefit from fast connections but is now opened to other colleges, K-12 schools and libraries. Because only educational and research sites are available on Internet2, it does not experience the congestion that often occurs on the commodity Internet. Although commercial sites are not on Internet2, Internet2 does provide a separate commodity Internet peering service. MOREnet participates in Internet2’s content peering service (CPS), which along with Qwest and Level3 gives MOREnet three connections to Internet for better redundancy and response.

What is BGP?

BGP, or the Border Gateway Protocol, is used between the network providers on the Internet to share routing information. If a provider has an IP network it wants others to reach it advertises the network to the other providers it connects to via BGP (its neighbors). These providers advertise the network to their neighbors and so on. Every route announcement contains the AS Number (see below) of each provider that has re-advertised it, in order. Because a router may have several different announcements for the same network from different neighbors it chooses the best route based on the number of AS numbers or "hops" in each announcement. BGP has a number of features to limit and control how routing information is propagated so that complex policies can be implemented.

What is an Autonomous System?

When speaking about BGP an AS or Autonomous System is a term used to refer to a network or set of networks under one administrative jurisdiction. Each AS is assigned a unique number (from 1 to 65000) to identify itself within the BGP protocol.

What is the difference between eBGP and iBGP? eBGP is used to speak between two providers in different autonomous systems (AS) and is pretty much what was described above. iBGP is used within a single AS to share routes around. It is only necessary if there is more than one router in the AS with external connectivity (which is common). It is necessary to have full BGP information at each router in the AS that is connected to more than one other router.

Multicast

What is IP Multicast?

Multicast allows a host to efficiently send a single IP packet to multiple destinations. The packet is replicated as necessary by routers throughout the network without ever having more than one copy of the packet traverse a particular network link. The sender doesn't need to do anything special or even know about all the designations. It simply sends the packet to a special IP address called a Multicast Group. Hosts that want to receive the packets send a special message called an IGMP Join to their nearest router and it handles the mechanics of requesting the stream from its upstream router. Multicast is primarily used for applications such as multimedia streaming where the data is broadcast at a steady rate and packet loss is acceptable.

NTP (network time protocol) and Session Directory Protocol advertisements are a few other examples of protocols that can use multicast. Multicast is attractive because it makes efficient use of network bandwidth. For a traditional RealAudio or other multimedia server to service 100 clients it must send 100 copies of the stream across its WAN link. With multicast each router involved in the broadcast replicates packets where necessary to make most efficient use of network links.

The drawback of multicast is that it is far more complex than unicast. Researchers have been working on the technology for more than a decade but it has only started to gain wide acceptance in the past few years. Most commercial providers still don't support it. MOREnet's support is still limited. Multicast saves bandwidth but requires routers to be smarter and keep track of more details (to have faster CPUs and more memory) than is required in a unicast environment.

What is a Multicast Group address?

Multicast groups uses a special range of IP addresses reserved for that purpose. These addresses range from 224.0.0.0 to 239.255.255.255.

What is PIM?

PIM stands for Protocol-Independent Multicast. The protocol was created by Cisco but has been adopted by several other vendors. PIM is a complex protocol that handles the transmission of multicast packets to all hosts in the multicast group while preventing loops and wasted bandwidth. PIM has two modes: dense mode and the newer sparse mode. PIM can manage some multicast groups in dense mode and some in sparse mode.

What is PIM Dense Mode?

Dense mode PIM is the older and simpler PIM mode. It works well in small networks where there are a large number of listeners, but is inefficient in larger networks. Cisco no longer recommends using dense mode. It works via a flood and prune strategy, where a multicast group is initially flooded to every router in the network, and then routers that don't have any listeners for a group send "prune" messages back upstream to block transmission. Every three minutes the routers begin flooding again and prunes must be repeated. A later addition to dense-mode called grafting allows routers to actively request a multicast group from the upstream router.

What is PIM Sparse Mode?

PIM sparse mode is more efficient when you have relatively few listeners for any given multicast stream. It saves bandwidth over Dense mode but is much more complex and requires a central router called a Rendezvous Point or RP to arbitrate traffic flow. When a receiver wants to receive traffic from a particular multicast group it sends a Join message toward the RP. This builds up a multicast path called the shared tree from the RP to the receiver as it travels toward the RP. When a source begins to transmit on a group the directly connected router initially encapsulates the packet in a Register message and sends it via Unicast to the RP. The RP sends the encapsulated multicast traffic down the shared tree to any listeners that have already joined. The RP also sends a join toward the sender to extend the shared tree to it. Now all traffic is flowing from sender to all receivers, although it is all passing through the RP rather than taking the most efficient path. If the bandwidth used by the session exceeds a certain threshold the router nearest each receiver can send a multicast join directly toward the sender. As the Join travels through the network a new multicast routing tree is created through the network, called the Shortest Path Tree. This is the most efficient path and once it is built the network will start using it to get from the sender to the receiver. This has just scratched the surface. See Cisco's website if you really want to know more.

What is MSDP?

The Multicast Source Discovery Protocol, or MSDP, is a protocol to allow multiple PIM sparse-mode domains to work together. Because PIM-SM requires a single RP router for each group it doesn't scale very well across administrative domains. With MSDP each domain has its own RPs and MSDP provides the conduit for the RPs to learn about each other's active multicast groups.

What is an RPF check?

Multicast routing adds a whole new complexity to normal unicast routing. The router doesn't just have to worry about which interfaces it must transmit the packets out on. It also has to consider the source IP of the packet and which interface it arrives on to prevent loops. It does this by performing an RPF, or Reverse Path Forwarding check. Normally this means looking up the source IP in the router's unicast routing table and seeing if it has a route to the source going out the interface that the packet came in on. If not then the packet didn't arrive on the most efficient path between source and the router and should be discarded. There are other criteria that RPF checks can be based on, but this gets complicated quickly.

Where can I learn more about Multicast?

The best place to learn about Multicast and how it relates to Cisco routers is at ftp://ftpeng.cisco.com/ipmulticast.html.

DNS

What is DNS?

DNS, the Domain Name Service, is a system for relating names to IP addresses. It is essentially a hierarchical, distributed database that spans thousands of computers across the globe. At the top of the hierarchy are the "root" name servers. All these servers know is the names and IP addresses of the name servers that serve each top-level domain (".com", ".edu", ".us", etc.). The name servers for each top-level domain know the IP addresses and name of name servers for the TLD's sub-domains. For instance, a name server for ".edu" can give you a list of the name servers for "mit.edu". The name servers for ".mit.edu" can tell you the IP addresses of servers within its domain, such as "www.mit.edu". It is also possible for it to create subdomains and assign these to still other name servers.

What is the difference between a primary and secondary DNS server?

From the point of view of the outside world there is no difference at all. The servers themselves have to know which is which, though. There is always only one primary DNS server for a domain. The secondaries are configured with the IP address of the primary server and they download all their information about the domain from it. Secondaries periodically check the primary for DNS changes and download them automatically when necessary.

Will MOREnet be the DNS server for my domain?

This is a service we provide to our members. We will act as either a primary server or a secondary. Many of our larger members want to run their own primary name server and that is fine with us. It gives them more control over their DNS entries and saves us some work. We will act as secondary if they request it. Many smaller sites lack people with the expertise to run a DNS server. We are happy to do this for them.
What DNS server should I configure my workstations to use?

If your site connects to MOREnet's Columbia hub then use these DNS servers:

* 150.199.1.10
* 150.199.8.1
* 150.199.101.1

If your site connects to the Jefferson City hub use these servers:

* 150.199.8.1
* 150.199.1.10
* 150.199.199.1

If your site connects to the Kansas City hub use these servers:

* 150.199.101.1
* 150.199.1.10
* 150.199.199.1

If your site connects to the Springfield hub use these servers:

* 150.199.199.1
* 150.199.8.1
* 150.199.101.1

If your site connects to the St. Louis hub use these servers:

* 150.199.178.1
* 150.199.8.1
* 150.199.1.10

What is the .US domain?

Each country has its own TLD (top-level domain). Many countries use theirs heavily, but many US citizens still prefer to use the generic top-level domains. The hierarchy of the .US domain is geographic. There is a subdomain for each state, such as "mo.us" or "ks.us". Beneath these are locality names such as "kansas-city.mo.us" or "springfield.mo.us". Anyone may register subdomains of locality domains. There also exist functional subdomains such as "k12.mo.us" for K12 schools, "cc.mo.us" for community colleges, "lib.mo.us" for libraries, "mus.mo.us" for museums, etc. The US domain registrar has until recently only had a few employees, so each of the locality and functional subdomains are run by volunteers. These volunteers are required to provide either free or low-cost registrations to those who request them.

What .US domains does MOREnet act as registrar for?

We provide service for the domains listed below. We operate these domains impartially and provide service to members and non-members alike. We don't allow any single entity to have more than one domain in the functional subdomains such as "k12.mo.us" or "cc.mo.us."

  • cc.mo.us gen.mo.us
  • k12.mo.us lib.mo.us
  • mus.mo.us tec.mo.us saint-louis.mo.us
  • st-louis.mo.us (deprecated; no new registrations are accepted)

How do I request a domain name?

Members should see the page on DNS services in Services. Non-members should e-mail register@more.net and fill out the form sent by the auto-responder.

Will you register domains for members in other domains such as .EDU or .ORG?

We used to provide this as a service to our members, but billing and tracking has become too much of a problem. If members register their own domains we will provide primary or secondary DNS service as needed. We would consider it polite if you e-mailed us about this before requesting the domain, though.

Network Management Tools

How do I use MRTG?

To view network data using MRTG, you must log in through MOREnet's free MyMOREnet application. MyMOREnet allows organizations to control who has access to their network data.

To use MRTG:

  1. Log in to MyMOREnet.
  2. Under Network on the Application Menu page, click MRTG Usage Statistics.
  3. To view a list of available reports for a router, click the name of the router.
  4. Click the report you wish to view.

The report displays graphs of network activity for the current day, week, month and year.

How often are the statistics in MRTG updated?

MRTG statistics are updated every five minutes.

What will NetFlow tell me?

MOREnet's NetFlow tool allows you to see which IP addresses at your site are passing the most traffic and which applications are being used most.

To view NetFlow reports:

  1. Log in to MOREnet's free MyMOREnet application. MyMOREnet allows organizations to control who has access to their network data.
  2. Under Network on the MyMOREnet Application Menu page, click NetFlow Reports.

For information on using Netflow Reports, in the blue NetFlow Reports menu on the right side of the page, click Help.

How can MOREnet's looking glass help me?

When troubleshooting network connectivity problems, it is often useful to run ping and trace route outside of your own network. The looking glass allows you to do either of these from one of MOREnet's core routers. Users can also view our BGP routing tables to troubleshoot the really hard problems. 

I can find my site, but there is no data or the data looks wrong. What is up?

If your site was recently activated or upgraded, it can take a few days for us to get all the network management tools set up for you. If correct information has not been available for three days or no changes have occurred recently, please e-mail help@more.net and we will look into the problem.