Internet user-bandwidth management and control tool

• US 6,085,241 A
• Filed: 07/22/1998
• Issued: 07/04/2000
• Est. Priority Date: 07/22/1998
• Status: Expired due to Term

• Current Assignee: Hanger Solutions, LLC (IP Investments Group LLC)
• Original Assignee: Amplify.net, Incorporated
• Inventors: Otis, Douglas A.
• Primary Examiner(s): VU, VIET DUY
• Application Number:
  US09/121,027
• Time in Patent Office: 713 Days
• Field of Search: 709/223, 709/224, 709/225, 709/226, 709/227, 709/229, 709/230, 709/232, 709/233, 709/235, 709/238, 709/250, 714/11, 714/47, 370/229, 370/230, 370/232, 370/234
• US Class Current: 709/223
• CPC Class Codes: H04L 41/5009 Determining service level performance parameters or violations of service level contracts, e.g. violations of agreed response time or mean time between failures [MTBF]
• CPC Class Codes: H04L 41/5067 Customer-centric QoS measurements
• CPC Class Codes: H04L 43/00 Arrangements for monitoring or testing data switching networks
• CPC Class Codes: H04L 43/045 for graphical visualisation of monitoring data
• CPC Class Codes: H04L 43/06 Generation of reports
• CPC Class Codes: H04L 43/067 using time frame reporting
• CPC Class Codes: H04L 43/0882 Utilisation of link capacity

Abstract

A bandwidth manager for network segments comprises a pair of media access controllers connected by respective MII data interfaces to IEEE 802.3 physical layer transceivers. A data selector connected amongst the media access controllers and physical layer transceivers allows the transceivers to be cross connected at their media-independent data interfaces while the media access controllers are simply used to monitor the symbols being exchanged. The data selector also allows the media access controllers and physical layer transceivers to be connected in pairs with each pair able to control the data traffic in each direction between two network segments. In such mode, bandwidth usage statistics are computed packet-by-packet for each IP-address under management and bandwidth excesses are either mitigated or blocked completely.

Claims

1. A network bandwidth management tool, comprising:

a first and a second network interface "A" and "B" that provide for a connection of a group of network users to network;

a first and a second physical layer interface transceiver respectively connected to the first and second network interfaces "A" and "B";

a first and a second media access controller (MAC) respectively connected by media-independent interfaces (MII) to receive symbol information from the first and second physical layer interface transceivers;

a cross-connection that patches through a receiver symbol output of the first physical layer interface transceiver to a transmitter symbol input of the second physical layer interface transceiver, and that patches through a receiver symbol output of the second physical layer interface transceiver to a transmitter symbol input of the first physical layer interface transceiver; and

a first and a second microcomputer respectively connected to the first and second MACs and that provide for a computing of traffic bandwidth associated with individual network addresses related to said group of users, wherein said computing of traffic bandwidth is based on symbol information read by each of the first and second MACs from said MII interfaces.

2. The tool of claim 1, further comprising:

a fail-safe mechanism providing for a direct interconnection of said group of network users to said network during an abnormal condition and for insertion of the tool via the first and second network interfaces "A" and "B" during a normal condition.

3. The tool of claim 2, wherein:

said normal and abnormal conditions are derived from a processor interrupt signal associated with either of the first and second MACs.

4. The tool of claim 1, further comprising:

a mode selector connected to disable the cross-connection and instead route said receiver symbol output of the first physical layer interface transceiver to a receiver symbol input of the first MAC, and route said receiver symbol output of the second physical layer interface transceiver to a receiver symbol input of the second MAC, and route said transmitter symbol input of the first physical layer interface transceiver to a transmitter symbol output of the second MAC, and route said transmitter symbol input of the second physical layer interface transceiver to a transmitter symbol output of the first MAC;

wherein data traffic moving from network interface "A" to "B" must be processed by the first MAC, and data traffic moving from network interface "B" to "A" must be processed by the second MAC.

5. The tool of claim 4, wherein:

the first and second microcomputers provide for a computing of traffic bandwidth statistics associated with individual network addresses related to said group of users, and said statistics are used on-the-fly to limit the rate at which individual data packets are passed between the network interfaces "A" and "B".

6. The tool of claim 5, wherein:

the first and second microcomputers provide independent time delays of individual packets to control bandwidth use based on static limits and dynamic-statistic information.

7. The tool of claim 5, wherein:

the first and second microcomputers provide for an independent time delay of individual data packets passed between the network interfaces "A" and "B" if a particular traffic bandwidth statistic associated with an individual network address indicates a rate limit would otherwise be exceeded.

8. The tool of claim 7, wherein:

the first and second microcomputers each maintain a ring buffer to spool and despool streams of data packets passed between the network interfaces "A" and "B".

9. The tool of claim 8, wherein:

the first and second microcomputers each include a mechanism to prevent hanging that would otherwise occur if the first or second MAC was currently acting as a bus master to control a main memory resource.

10. A network bandwidth manager, comprising:

a first and a second ETHERNET network interface "A" and "B" that provide for an internal protocol (IP) connection of a group of network users to network;

a first and a second physical layer interface transceiver respectively connected to the first and second network interfaces "A" and "B";

a first and a second media access controller (MAC) respectively connected by media-independent interfaces (MII) to receive symbol information from the first and second physical layer interface transceivers;

a cross-connection that patches through a receiver symbol output of the first physical layer interface transceiver to a transmitter symbol input of the second physical layer interface transceiver, and that patches through a receiver symbol output of the second physical layer interface transceiver to a transmitter symbol input of the first physical layer interface transceiver;

a first and a second microcomputer respectively connected to the first and second MACs and that provide for a computing of traffic bandwidth associated with individual network IP-addresses related to said group of users, wherein said computing of traffic bandwidth is based on symbol information read by each of the first and second MACs from said MII interfaces;

a fail-safe mechanism providing for a direct interconnection of said group of network users to said network during an abnormal condition and for insertion of the tool via the first and second network interfaces "A" and "B" during a normal condition, wherein, said normal and abnormal conditions are derived from a processor interrupt signal associated with either of the first and second MACs; and

a mode selector connected to disable the crossconnection and instead route said receiver symbol output of the first physical layer interface transceiver to a receiver symbol input of the first MAC, and route said receiver symbol output of the second physical layer interface transceiver to a receiver symbol input of the second MAC, and route said transmitter symbol input of the first physical layer interface transceiver to a transmitter symbol output of the second MAC, and route said transmitter symbol input of the second physical layer interface transceiver to a transmitter symbol output of the first MAC, wherein, data traffic moving from network interface "A" to "B" must be processed by the first MAC, and data traffic moving from network interface "B" to "A" must be processed by the second MAC.

11. The bandwidth manager of claim 10, wherein:

the first and second microcomputers compute traffic bandwidth statistics associated with individual network IP-addresses related to said group of users, and said statistics are used on-the-fly to limit the rate at which individual data packets can be passed between the network interfaces "A" and "B";

the first and second microcomputers provide for an independent time delay of individual data packets passed between the network interfaces "A" and "B" if a particular traffic bandwidth statistic associated with an individual network IP-address indicates a rate limit would otherwise be exceeded;

the first and second microcomputers each maintain a ring buffer to spool and despool streams of data packets passed between the network interfaces "A" and "B"; and

the first and second microcomputers each include a processor-interrupt mechanism to prevent hanging that would otherwise occur if the first or second MAC was currently acting as a bus master to control a mains memory resource when the first and second microcomputers attempted a main memory access.

12. The bandwidth manager of claim 11, wherein:

the first and second microcomputers automatically adjust a connection rate bandwidth to take advantage of any remaining connection capacity and maintain a pre-determined margin within a pre-determined connection bandwidth size.

13. The bandwidth manager of claim 12, wherein:

the first and second microcomputers automatically adjust said connection rate bandwidth by controlling a direction of bandwidth slew and simultaneously adjusting an overall maximum bandwidth for any one particular connection;

wherein, if said pre-determined margin is inadequate, an allowable top limit for said connection rate is slowly reduced over time and eventually lowers to a lowest level if inactive for a particular time, and such that a new exchange must build up said connection rate bandwidth again.

14. The bandwidth manager of claim 13, further comprising:

a limiter to control the maximum duty cycle that can be elicited by a connection client, wherein a statistic is integrated over a long term connection history for said connection client, and such statistic is later used as a dominate control to limit bandwidth to said connection client.

15. The bandwidth manager of claim 14, further comprising:

means for limiting maximum bandwidth allocations to particular connections that maintain an excessive connection bandwidth over a prolonged period such that a single connection cannot abuse the overall connection bandwidth of the system to the chagrin of other connections; and

system notification means to notify the system when any uncontrolled bandwidth protocols or abuse are detected, wherein such information is returned as a dropped-packet count ear-marked for particular connections to protect a connection where rejected packets could otherwise consume valuable resources.