Method, node, and system for controlling network traffic based on token ring
A method for controlling network traffic based on a token ring includes: acquiring, by a node that obtains a token, network congestion situation information carried in the token, where the network congestion situation information includes the total number of to-be-retransmitted messages requested by nodes in a current token passing period; comparing the total number of to-be-retransmitted messages requested by the nodes in the current token passing period with a congestion threshold; determining according to a comparison result, the number of messages that can be transmitted; and transmitting messages according to the number. Because the total number of to-be-retransmitted messages requested by nodes in a current token passing period is added into a token to reflect a network congestion situation, a node is capable of reducing the number of messages to be transmitted when a network is severely congested, thereby reducing network load and improving cluster availability.
1. A method for controlling network traffic based on a token ring, comprising:
acquiring, by a node that obtains a token, network congestion situation information carried in the token, wherein the network congestion situation information comprises the total number of to-be-retransmitted messages requested by nodes in a current token passing period;
comparing, by the node that obtains the token, a total number of to-be-retransmitted messages requested by the nodes in the current token passing period with a congestion threshold, wherein the congestion threshold is determined according to a network congestion test;
determining, by the node that obtains the token and according to a comparison result, a number of messages that can be transmitted, wherein the node determines the number of messages that can be transmitted based on whether the total number of to-be-retransmitted messages is less than or greater than or equal to the congestion threshold; and
transmitting messages according to the number of messages that can be transmitted,
wherein determining, by the node that obtains the token and according to the comparison result, the number of messages that can be transmitted, comprises:
when the total number of to-be-retransmitted messages requested by the nodes in the current token passing period is less than the congestion threshold, transmitting, by the node that obtains the token, messages according to a determined first number of messages that can be transmitted, wherein the determined first number of messages that can be transmitted is determined based on a maximum number of messages that each node can transmit after receiving the token and a total number of messages that have been transmitted by all nodes that have transmitted messages in the current token passing period; and
when the total number of to-be-retransmitted messages requested by the nodes in the current token passing period is greater than or equal to the congestion threshold, transmitting, by the node that obtains the token, messages according to a determined second number of messages that can be transmitted, wherein the second number of messages that can be transmitted is determined by the network congestion situation information and the first number of messages that can be transmitted, and the second number of messages that can be transmitted is less than the first number of messages that can be transmitted.
2. The method according to claim 1 , further comprising determining, by the node that obtains the token and according to a local receive buffer queue, the number of messages that are locally requested to be retransmitted, and updating, according to the number of messages that are locally requested to be retransmitted, the total number of to-be-retransmitted messages requested by the nodes in the current token passing period.
3. The method according to claim 1 , wherein the network congestion situation information further comprises:
a number of times of congestion occurrences; and
the second number of messages that can be
transmitted
=
N
×
1
2
M
,
or the second number of messages that can be
transmitted
=
N
×
1
2
M
+
1
,
wherein N is the first number of messages that can be transmitted, and wherein M is the times of congestion occurrences.
4. The method according to claim 1 , wherein a times of congestion occurrences is configured on the node that obtains the token, wherein the second number of messages that can be
transmitted
=
N
×
1
2
M
,
or the second number of messages that can be
transmitted
=
N
×
1
2
M
+
1
,
wherein N is the first number of messages that can be transmitted, and wherein M is the times of congestion occurrences.
5. The method according to claim 3 , further comprising:
when the total number of to-be-retransmitted messages requested by the nodes in the current token passing period is less than the congestion threshold and the times of congestion occurrences is not zero, resetting, by the node that obtains the token, the times of congestion occurrences to zero; and
when the total number of to-be-retransmitted messages requested by the nodes in the current token passing period is greater than or equal to the congestion threshold, increasing, by the node that obtains the token, the times of congestion occurrences by one.
6. The method according to claim 4 , further comprising:
when the total number of to-be-retransmitted messages requested by the nodes in the current token passing period is less than the congestion threshold and the times of congestion occurrences is not zero, resetting, by the node that obtains the token, the times of congestion occurrences to zero; and
when the total number of to-be-retransmitted messages requested by the nodes in the current token passing period is greater than or equal to the congestion threshold, increasing, by the node that obtains the token, the times of congestion occurrences by one.
7. A token ring node, comprising:
a computer processor configured to:
acquire network congestion situation information carried in a token, wherein the network congestion situation information comprises a total number of to-be-retransmitted messages requested by nodes in a current token passing period; and
compare the total number of to-be-retransmitted messages requested by the nodes in the current token passing period with a congestion threshold, determine, according to a comparison result, a number of messages that can be transmitted, and transmit messages according to the number of messages that can be transmitted, wherein the congestion threshold is determined according to a network congestion test, and wherein the number of messages that can be transmitted is determined based on whether the total number of to-be-retransmitted messages is less than or greater than or equal to the congestion threshold;
when the total number of to-be-retransmitted messages requested by the nodes in the current token passing period is less than the congestion threshold, transmit messages according to a determined first number of messages that can be transmitted, wherein the determined first number of messages that can be transmitted is determined based on a maximum number of messages that each node can transmit after receiving the token and a total number of messages that have been transmitted by all nodes that have transmitted messages in the current token passing period; and
when the total number of to-be-retransmitted messages requested by the nodes in the current token passing period is greater than or equal to the congestion threshold, transmit messages according to a determined second number of messages that can be transmitted,
wherein the second number of messages that can be transmitted is determined by the network congestion situation information and the first number of messages that can be transmitted, and
wherein the second number of messages that can be transmitted is less than the first number of messages that can be transmitted.
8. The token ring node according to claim 7 , wherein the computer processor is configured to determine, according to a local receive buffer queue, the number of messages that are locally requested to be retransmitted, and update, according to the number of messages that are locally requested to be retransmitted, the total number of to-be-retransmitted messages requested by the nodes in the current token passing period.
9. A token ring node, comprising:
a computer processor configured to:
acquire network congestion situation information carried in a token, wherein the network congestion situation information comprises a total number of to-be-retransmitted messages requested by nodes in a current token passing period; and
compare the total number of to-be-retransmitted messages requested by the nodes in the current token passing period with a congestion threshold, determine, according to a comparison result, a number of messages that can be transmitted, and transmit messages according to the number of messages that can be transmitted, wherein the congestion threshold is determined according to a network congestion test, and wherein the number of messages that can be transmitted is determined based on whether the total number of to-be-retransmitted messages is less than or greater than or equal to the congestion threshold,
wherein the network congestion situation information further comprises the times of congestion occurrences, or the times of congestion occurrences is configured on a node that obtains the token, and
wherein the computer processor is further configured to:
when the total number of to-be-retransmitted messages requested by the nodes in the current token passing period is less than the congestion threshold and the times of congestion occurrences is not zero, reset the times of congestion occurrences to zero; and
when the total number of to-be-retransmitted messages requested by the nodes in the current token passing period is greater than or equal to the congestion threshold, increase the times of congestion occurrences by one.
10. A system for controlling network traffic based on a token ring, comprising:
a ring network consisting of at least three nodes, wherein each of the nodes passes a token in a predefined order; and
a node that obtains the token configured to:
acquire network congestion situation information carried in the token, wherein the network congestion situation information comprises a total number of to-be-retransmitted messages requested by the nodes in a current token passing period;
compare the total number of to-be-retransmitted messages requested by the nodes in the current token passing period with a congestion threshold, wherein the congestion threshold is determined according to a network congestion test;
determine, according to a comparison result, a number of messages that can be transmitted, wherein the number of messages that can be transmitted is determined based on whether the total number of to-be-retransmitted messages is less than or greater than or equal to the congestion threshold; and
transmit messages according to the number of messages that can be transmitted,
wherein the node that obtains the token is configured to:
when the total number of to-be-retransmitted messages requested by the nodes in the current token passing period is less than the congestion threshold, transmit messages according to a determined first number of messages that can be transmitted, wherein the determined first number of messages that can be transmitted is determined based on a maximum number of messages that each node can transmit after receiving the token and a total number of messages that have been transmitted by all nodes that have transmitted messages in the current token passing period; and
when the total number of to-be-retransmitted messages requested by the nodes in the current token passing period is greater than or equal to the congestion threshold, transmit messages according to a determined second number of messages that can be transmitted,
wherein the second number of messages that can be transmitted is determined by the network congestion situation information and the first number of messages that can be transmitted, and
wherein the second number of messages that can be transmitted is less than the first number of messages that can be transmitted.
11. The system according to claim 10 , wherein the node that obtains the token is further configured to determine, according to a local receive buffer queue, the number of messages that are locally requested to be retransmitted, and update, according to the number of messages that are locally requested to be retransmitted, the total number of to-be-retransmitted messages requested by the nodes in the current token passing period.
12. A system for controlling network traffic based on a token ring, comprising:
a ring network consisting of at least three nodes, wherein each of the nodes passes a token in a predefined order; and
a node that obtains the token configured to:
acquire network congestion situation information carried in the token, wherein the network congestion situation information comprises a total number of to-be-retransmitted messages requested by the nodes in a current token passing period;
compare the total number of to-be-retransmitted messages requested by the nodes in the current token passing period with a congestion threshold, wherein the congestion threshold is determined according to a network congestion test;
determine, according to a comparison result, a number of messages that can be transmitted, wherein the number of messages that can be transmitted is determined based on whether the total number of to-be-retransmitted messages is less than or greater than or equal to the congestion threshold; and
transmit messages according to the number of messages that can be transmitted,
wherein the network congestion situation information further comprises the times of congestion occurrences, or the times of congestion occurrences is configured on the node that obtains the token, and
wherein the node is further configured to:
when the total number of to-be-retransmitted messages requested by the nodes in the current token passing period is less than the congestion threshold and the times of congestion occurrences is not zero, reset the times of congestion occurrences to zero; and
when the total number of to-be-retransmitted messages requested by the nodes in the current token passing period is greater than or equal to the congestion threshold, increase the times of congestion occurrences by one.