REAL-TIME MONITORING OF A ROUTING SERVER

• US 20090172152A1
• Filed: 12/31/2007
• Published: 07/02/2009
• Est. Priority Date: 12/31/2007
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Patiejunas, Kestutis
• Granted Patent: US 7,818,422 B2
• Time in Patent Office: Days
• US Class Current: 709/224
• CPC Class Codes: H04L 43/0864 Round trip delays
• CPC Class Codes: H04L 45/26 Route discovery packet

Abstract

Calculating the roundtrip delay between first and second computers due to the latency of a routing server. The first computer receives an incoming message from the second computer via the routing server. The incoming message includes incoming time information indicative of processing time at the second computer and outgoing time information indicative of a time when a first outgoing message was sent by the first computer. The first computer calculates an observed roundtrip time indicating the routing server latency for a first outgoing message and the incoming message as a function of a time when the first computer received the incoming message, the outgoing time information, and the incoming time information.

Claims

1. A method for determining the round trip latency between first and second computers due to a routing server therebetween, said method comprising:

sending a first outgoing message from the first computer to the second computer via the routing server, said first outgoing message including outgoing time information indicative of a time when the first outgoing message was sent by the first computer according to a first clock associated with the first computer;

receiving an incoming message from the second computer to the first computer via the routing server, said incoming message being sent by the second computer in response to receipt of the first outgoing message sent by the first computer, said incoming message including incoming time information indicative of a time when the incoming message was sent by the second computer according to a second clock associated with the second computer, said incoming time information further indicating processing time at the second computer, and said incoming message further including the outgoing time information;

determining a time when the first computer receives the incoming message according to the first clock; and

calculating a first observed roundtrip time indicating the routing server latency for the outgoing and incoming messages as a function of the determined time, the outgoing time information, and the incoming time information, wherein said first observed roundtrip time is equal to the difference between the time when the first outgoing message was sent according to the first clock and the time the incoming message was received according to the first clock, minus the processing time at the second computer as indicated by the incoming time information.

2. The method of claim 1, wherein the first computer is a remote desktop client, the second computer is a remote desktop server, the routing server is a remote desktop routing server, and the outgoing and incoming messages are in a remote desktop protocol format.

3. The method of claim 1, further comprising providing the outgoing time information in a header of the first outgoing message and providing the incoming time information in a header of the incoming message.

4. The method of claim 1, further comprising encrypting the first outgoing message and the incoming message, and wherein the outgoing time information and the incoming time information are not encrypted.

5. The method of claim 1, wherein the first clock and second clock are not synchronized.

6. The method of claim 1, further comprising sending a second outgoing message from the first computer to the second computer in response to receipt of the incoming message from the second computer, said second outgoing message including the outgoing time information further indicating a time the received incoming message was sent according to the second clock and a processing time at the first computer;

and

wherein, in response to receiving the second outgoing message, the second computer determines a time when the second computer receives the second outgoing message according to the second clock and the second computer calculates a second observed roundtrip time indicating the routing sever latency for the incoming message and the second outgoing message as a function of the determined time, the outgoing time information, and the incoming time information, wherein said second observed roundtrip time is equal to the difference between the time when the second outgoing message was sent according to the first clock and the time the incoming message was received according to the second clock, minus the processing time at the first computer as indicated by the incoming time information.

7. The method of claim 6, wherein the routing server accesses the second outgoing message and further comprising calculating a performance metric according to the accessed first observed roundtrip time.

8. The method of claim 7, further comprising refusing, by the routing server, connections from other computers if the accessed first observed roundtrip time is above a threshold and allowing additional connections from other computers if the accessed first observed roundtrip time is below the threshold.

10. The method of claim 1, wherein the first computer is a remote desktop client, the second computer is a remote desktop server, the plurality of routing servers are remote desktop routing servers, and the first and second outgoing messages and the incoming message are in a remote desktop protocol format.

11. The method of claim 1, further comprising providing the outgoing time information in a header of the first and second outgoing messages and providing the incoming time information in a header of the incoming message.

12. The method of claim 1, further comprising encrypting the first and second outgoing messages and the incoming message, and wherein the outgoing time information and the incoming time information are not encrypted.

13. The method of claim 1, wherein the first clock and second clock are not synchronized.

9. A method for selecting a routing server from a plurality of available routing servers by a first computer, said first computer sending outgoing messages to a second computer and receiving incoming messages from said second computer, said first computer having a first clock associated therewith and said second computer having a second clock associated therewith, said incoming and outgoing messages being sent via at least one of the plurality of routing servers, said method comprising:

receiving an incoming message from the second computer to the first computer via a first routing server, said first routing server being one of the plurality of available routing servers, said incoming message being sent by the second computer in response to receipt of a first outgoing message sent by the first computer, said incoming message including incoming time information indicative of a time when the incoming message was sent by the second computer according to the second clock associated with the second computer, said incoming time information further indicating processing time at the second computer, and said incoming message further including the outgoing time information, said outgoing time information being indicative of a time when said first outgoing message was sent by the first computer according to the first clock associated with the first computer;

determining a time when the first computer receives the incoming message according to the first clock;

calculating a first observed roundtrip time indicating the first routing server latency for the outgoing and incoming messages as a function of the determined time, the outgoing time information, and the incoming time information, wherein said first observed roundtrip time is equal to the difference between the time when the first outgoing message was sent according to the first clock and the time the incoming message was received according to the first clock, minus the processing time at the second computer as indicated by the incoming time information; and

wherein;

if the first observed roundtrip time is below a threshold, the first computer sends a second outgoing message to the second computer via the first routing server, said threshold indicating an acceptable response time between the first computer and second computer; and

if the first observed roundtrip time is above the threshold, the first computer sends the second outgoing message to the second computer via a second routing server selected from the plurality of available routing servers, said second routing server being different than the first routing server.

14. A computer-readable storage medium including a latency data structure of a second outgoing message for tracking an observed roundtrip time between a remote desktop client and a remote desktop server via a first remote desktop routing server based on a first outgoing message and incoming message, said first outgoing message, said incoming message, and said second outgoing message being in a remote desktop protocol format, said first remote desktop routing server being one of a plurality of available remote desktop routing servers, said latency data structure comprising:

a sender time for storing a time the second outgoing message was sent by the remote desktop client according to a first clock associated with the remote desktop client;

a processing delta time for storing a remote desktop client processing time, said remote desktop client processing time equal to the difference between a time the incoming message arrived according to the first clock and the time the second outgoing message was sent according to the remote desktop client clock;

a peer time for storing a time the incoming message was sent according to a second clock, said time the incoming message was sent being included in the incoming message; and

an observed propagation time for storing a calculated observed roundtrip time for the first remote desktop routing server, wherein in response to receiving the incoming message, the remote desktop client calculates the observed roundtrip time, said observed roundtrip time being equal to the difference between a time the first outgoing message was sent and the time the incoming message was received, minus a remote desktop server processing time, said time the first outgoing message was sent and said remote desktop server processing time included in the incoming message.

15. The computer-readable storage medium of claim 14, wherein:

the remote desktop client sends the second outgoing message to the remote desktop server via the first desktop routing server if the first observed roundtrip time is below a threshold, said threshold indicating an acceptable response time between the remote desktop client and remote desktop server; and

the remote desktop client sends the second outgoing message to the remote desktop server via a second desktop routing server if the first observed roundtrip time is above the threshold, said second desktop routing server being one of the plurality of available desktop routing servers, said second desktop routing server being different than the first desktop routing server.

16. The computer-readable storage medium of claim 14, wherein the latency data structure is included in a header of the second outgoing message.

17. The computer-readable storage medium of claim 14, wherein the latency data structure is not encrypted and the second outgoing message is encrypted.

18. The computer-readable storage medium of claim 14, wherein the remote desktop client clock and remote desktop server clock are not synchronized.

19. The computer-readable storage medium of claim 14, wherein the first remote desktop routing server accesses the observed propagation time of the latency data structure and wherein the first remote desktop routing server allows additional connections from additional remote desktop clients based on the accessed observed propagation time.

20. The computer-readable storage medium of claim 14, wherein the first remote desktop routing server accesses the observed propagation time of the latency data structure and calculates a performance metric based on the accessed observed propagation time.