Movatterモバイル変換


[0]ホーム

URL:


CN103763215A - Chip array priority mapping method and system - Google Patents

Chip array priority mapping method and system
Download PDF

Info

Publication number
CN103763215A
CN103763215ACN201410011437.5ACN201410011437ACN103763215ACN 103763215 ACN103763215 ACN 103763215ACN 201410011437 ACN201410011437 ACN 201410011437ACN 103763215 ACN103763215 ACN 103763215A
Authority
CN
China
Prior art keywords
priority
message
mapping
chip
queue
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201410011437.5A
Other languages
Chinese (zh)
Inventor
吴瑞
张梅
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Maipu Communication Technology Co Ltd
Original Assignee
Maipu Communication Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Maipu Communication Technology Co LtdfiledCriticalMaipu Communication Technology Co Ltd
Priority to CN201410011437.5ApriorityCriticalpatent/CN103763215A/en
Publication of CN103763215ApublicationCriticalpatent/CN103763215A/en
Pendinglegal-statusCriticalCurrent

Links

Images

Landscapes

Abstract

The invention relates to the field of network communication and discloses a chip array priority mapping method. Through the method, priority forwarding of control management messages in a virtual switching link is effectively achieved. The method includes the first step of identifying received messages and distinguishing service messages and control messages, the second step of remapping the highest-priority service messages to other priorities, the third step of mapping the priority of the control messages to the highest priority, the fourth step of repackaging message descriptors and identifying the priority of the current messages again, and the fifth step of allowing the virtual switching link to forward the messages according to the new priority. The invention further discloses a chip array priority mapping system. The chip array priority mapping system comprises a message identification module, a message mapping module, a message packaging module and a message switching module. The chip array priority mapping method and system are not restricted to inherent limitation of hardware, and firmware input, manpower cost and time cost are effectively reduced. The chip array priority mapping method and system can be widely suitable for chips with a limit of the array number to establish a virtual switching system.

Description

Chip queue priority mapping method and system
Technical field
The present invention relates to network communication field, relate in particular to chip queue priority mapping method and the system in virtual exchange system design, used.
Background technology
In virtual switch technological system building process, because control and management message (or referred to as the controlling message) kind of virtual switch link transmission is a lot, and relate to the packet sending and receiving of striding equipment application protocol.For guaranteeing the stability of virtual exchange system, must guarantee to control message can reliably be forwarded, conventionally namely make the type message forward by highest-priority queue, otherwise when impacting, service message makes to control after message dropping, can cause concussion or the division of virtual exchange system, affect normal data and forward.When the hardware chip using has number of queues restriction, this problem will certainly exist.As conventional exchange chip, chip queue only has altogether 8, corresponding 8 different priority, and each agreement and data message use have been distributed in 8 queues.
In order to ensure control and management message, can preferentially be forwarded, prior art generally all can only guarantee by using method, and the total number of restriction message can not exceed or approach the bandwidth of virtual switch link.This mode to the stability of system be do not have guaranteed, and because virtual switch technology is to use the resilient channels of switching engine to communicate, in Software for Design, not realizing at present the different Q OS(service quality to resilient channels) value carries out data statistics, when the total number of message exceedes or approach the bandwidth of virtual switch link, will inevitably there is the problem that control and management message is dropped, cause the virtual exchange system in actual application unstable, in the serious situation of the limit, cause virtual exchange system division, affecting Network normally moves.
Summary of the invention
The object of the invention is, a kind of chip queue priority mapping method is provided, effectively solve the control and management message prior forwarding problems in virtual switch link.
The present invention solve the technical problem, and the technical scheme of employing is that chip queue priority mapping method, comprises step:
A, the message of receiving is identified differentiated service message and control message;
B, the service message of limit priority is remapped to other priority;
C, will to control message priority-level mapping be limit priority;
The lay equal stress on priority of the current message of new logo of d, Reseal message descriptor;
E, virtual switch link E-Packet according to new priority.
Concrete, the QOS grade that described priority is message.
More specifically, described QOS grade equates with chip number of queues.
Preferably, steps d is specially, and Reseal message descriptor is also encapsulated into HIGIG protocol headers by the message priority remapping.
Recommend, step b is specially:
B1, non-lowest priority service message is remapped to low one-level priority;
B2, constant to lowest priority service message maintenance priority.
Another object of the present invention is, a kind of chip queue priority mapped system is provided, and comprises message identification module, message mapping block, message package module and message switching module;
Message identification module, identifies the message of receiving, differentiated service message and control message;
Message mapping block, is remapped to other priority by the service message of limit priority; To control message priority-level mapping is limit priority;
Message package module, the lay equal stress on priority of the current message of new logo of Reseal message descriptor;
Message switching module, E-Packets according to new priority.
Concrete, the QOS grade that described priority is message.
More specifically, described QOS grade equates with chip number of queues.
Preferably, described message mapping block, further Reseal message descriptor is also encapsulated into HIGIG protocol headers by the message priority remapping.
Recommend, described message mapping block is further remapped to low one-level priority to non-lowest priority service message; To lowest priority service message, keep priority constant.
The invention has the beneficial effects as follows, be not limited to the inherent limitations of hardware, effectively reduce firmware and drop into and manpower time cost.The present invention is as the basic module of system, and End-Customer end can't perceive, and the chip that can be widely used in number of queues restriction carries out the structure of virtual exchange system.
Accompanying drawing explanation
Fig. 1 is the flow chart of embodiment;
Fig. 2 is system configuration schematic diagram of the present invention.
Embodiment
Below in conjunction with drawings and Examples, describe technical scheme of the present invention in detail.
The present invention has utilized message to enter after chip, revises the method that priority remaps, and guarantees that the control message prior of queue forwards, and finally realizes queue compression.
Embodiment
This example is take the exchange chip commonly used as example, the QOS queue one of this chip has 8 queues, virtual switch port also only has 8 QOS queues, be respectively QOS0~7, corresponding 8 QOS grades and 8 priority, with 8 queues of user be overlapping, the control message that does not have independent queue to leave virtual switch for is used, acquiescence QOS queue and internal priority also will participate in the conversion operations of 802.1p, TOS, DSCP.As shown in Figure 1, concrete steps comprise chip queue priority mapping method flow process:
When port, receiving priority is the corresponding limit priority of 7() message after, chip can be identified by the type of message.If control message, hold it in QOS7 queue forwarding, if service message also can enter QOS7 queue according to the original priority of this message, and then the internal mapping table having defined by comparison, by message descriptor of this message Reseal, the priority of describing this service message is low one-level 6, send into 6 queues of virtual switch port-qos, virtual switch port can divest the message descriptor of this message, afterwards the priority 6 after mapping is write to HIGIG protocol headers and encapsulate this message, this message will be according to this priority at virtual switch link repeating.In repeating process, service message compress mode adopts original message priority except QOS0(correspondence lowest priority) the mode of one-level is fallen, be that QOS7 queue messages drops to QOS6, QOS6 queue messages drops to QOS5, by that analogy, until QOS1 queue, QOS0 queue messages priority remains unchanged.Thus by 8 priority mapping to 7 queues of service message, for the reserved limit priority QOS7 queue of control and management message of virtual switch link, the control message that port is received, can be mapped to highest-priority queue or remain on highest-priority queue, finally realizing queue compression.
The bandwidth of supposing this routine virtual switch link is 1G, by pouring into summation, exceedes virtual switch link bandwidth 5% different Q OS value and tests.The QOS value of these two flows is respectively QOS7 and QOS6, and they send respectively and account for bandwidth is 10% and 95% known unicast message flow.Finally check the packet receiving result after virtual switch link forwards.Result is that the message of QOS7 queue still can be received 10% flow, there will not be and substitute.The message of QOS6 queue can be received 90% flow, loses 5% flow, strict guarantee is described the message of high priority preferentially forwarded.When two streams of replacing are respectively QOS1 and QOS0, result is respectively to lose 2.5% flow, and this is because the message of QOS1 and QOS0 is all compressed to QOS0 queue.The queue of checking most critical is QOS7 queue, and by filling, wrap test is the transmission that can not affect the control message of QOS7 queue in QOS6 queue excess load, and now virtual exchange system, in stable state, can not be subject to the impact of excess load.Can find out, this spr chip queue priority mapping method, the platoon ratio being compressed is average, and duty ratio is more balanced, can not produce special impact to the queue of certain priority.If by the service message of limit priority, be fixedly mapped to other priority queries, although program is simpler, may produce larger impact to these other priority queries, be unfavorable for the equilibrium of load.
Chip queue priority mapped system structure of the present invention, referring to Fig. 2, comprises message identification module, message mapping block, message package module and message switching module.
Message identification module, identifies the message of receiving, differentiated service message and control message.
Message mapping block, is remapped to other priority by the service message of limit priority; To control message priority-level mapping is limit priority.Here the QOS grade that priority is message also equates with chip number of queues.
Message mapping block, further Reseal message descriptor is also encapsulated into Higig head by the message priority remapping.Non-lowest priority service message is remapped to low one-level priority; To lowest priority service message, keep priority constant.
Message package module, the lay equal stress on priority of the current message of new logo of Reseal message descriptor.
Message switching module, E-Packets according to new priority.

Claims (10)

CN201410011437.5A2014-01-102014-01-10Chip array priority mapping method and systemPendingCN103763215A (en)

Priority Applications (1)

Application NumberPriority DateFiling DateTitle
CN201410011437.5ACN103763215A (en)2014-01-102014-01-10Chip array priority mapping method and system

Applications Claiming Priority (1)

Application NumberPriority DateFiling DateTitle
CN201410011437.5ACN103763215A (en)2014-01-102014-01-10Chip array priority mapping method and system

Publications (1)

Publication NumberPublication Date
CN103763215Atrue CN103763215A (en)2014-04-30

Family

ID=50530377

Family Applications (1)

Application NumberTitlePriority DateFiling Date
CN201410011437.5APendingCN103763215A (en)2014-01-102014-01-10Chip array priority mapping method and system

Country Status (1)

CountryLink
CN (1)CN103763215A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN109167739A (en)*2018-09-192019-01-08盛科网络(苏州)有限公司Message processing method and system in a kind of switchboard stacked system
CN114415969A (en)*2022-02-092022-04-29杭州云合智网技术有限公司Dynamic storage method for message of switching chip
CN117793021A (en)*2023-12-262024-03-29无锡众星微系统技术有限公司Sideband management message processing method and circuit

Citations (6)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN1758625A (en)*2004-10-092006-04-12华为技术有限公司Method for classification processing message
CN1913486A (en)*2005-08-102007-02-14中兴通讯股份有限公司Method and device for strengthening safety of protocol message
CN101163112A (en)*2007-11-152008-04-16福建星网锐捷网络有限公司Packet processing method in switchboard stacked system and switchboard equipment
CN101980489A (en)*2010-10-282011-02-23中兴通讯股份有限公司Protection method and system for preventing protocol message from attacking CPU
CN102195858A (en)*2010-03-182011-09-21武汉迈威光电技术有限公司Overload protection technique of carrier Ethernet ring network
CN102377652A (en)*2010-08-172012-03-14盛科网络(苏州)有限公司Method for protecting message of important protocol by using chip

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN1758625A (en)*2004-10-092006-04-12华为技术有限公司Method for classification processing message
CN1913486A (en)*2005-08-102007-02-14中兴通讯股份有限公司Method and device for strengthening safety of protocol message
CN101163112A (en)*2007-11-152008-04-16福建星网锐捷网络有限公司Packet processing method in switchboard stacked system and switchboard equipment
CN102195858A (en)*2010-03-182011-09-21武汉迈威光电技术有限公司Overload protection technique of carrier Ethernet ring network
CN102377652A (en)*2010-08-172012-03-14盛科网络(苏州)有限公司Method for protecting message of important protocol by using chip
CN101980489A (en)*2010-10-282011-02-23中兴通讯股份有限公司Protection method and system for preventing protocol message from attacking CPU

Cited By (4)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN109167739A (en)*2018-09-192019-01-08盛科网络(苏州)有限公司Message processing method and system in a kind of switchboard stacked system
CN114415969A (en)*2022-02-092022-04-29杭州云合智网技术有限公司Dynamic storage method for message of switching chip
CN114415969B (en)*2022-02-092023-09-29杭州云合智网技术有限公司Method for dynamically storing messages of exchange chip
CN117793021A (en)*2023-12-262024-03-29无锡众星微系统技术有限公司Sideband management message processing method and circuit

Similar Documents

PublicationPublication DateTitle
US10084716B2 (en)Flexible application of congestion control measures
US10644800B2 (en)Channel bonding techniques in a network
KR101504144B1 (en)System and method for multi-channel packet transmission
US7969882B2 (en)Packet transfer rate monitoring control apparatus, method, and program
US20220078123A1 (en)Methods and apparatus for packet dropping in a fronthaul network
JP5347836B2 (en) Communication apparatus and communication method
CN1881979B (en)Ethernet physical layer low-speed transmission realizing method and its applied network apparatus
CN106385386A (en)Application servo-actuated intelligent floe control method
WO2020063339A1 (en)Method, device and system for realizing data transmission
EP2903226A1 (en)Link aggregation method and device for traffic distribution
CN110868359B (en)Network congestion control method
CN114124830B (en) A RDMA service quality assurance method and system for multiple application scenarios in data centers
US10771872B2 (en)Traffic management with limited-buffer switches
CN103763215A (en)Chip array priority mapping method and system
CN104184679A (en)Method of scheduling message in intelligent substation and device
CN109995608B (en)Network rate calculation method and device
CN101212467A (en)MPLS network service scheduling method
CN103516619B (en)The method and device of Bandwidth adjustment in a kind of network virtualization system
CN102932272A (en)Message forwarding method and forwarding device based on flow capacity
CN113038530B (en)High-efficiency transmission method for packet service of QoS guarantee of satellite mobile communication system
CN116192771B (en) A priority queue management method and system
US11870708B2 (en)Congestion control method and apparatus
US8213306B1 (en)Apparatus and methods for aggregation of communication bandwidth over multiple wireless communication links
CN104283647A (en)Data transmission method and device based on air interface and network communication device
CN105282053A (en)Method and system for preventing output head-of-line of network exchange chip from blocking

Legal Events

DateCodeTitleDescription
C06Publication
PB01Publication
C10Entry into substantive examination
SE01Entry into force of request for substantive examination
RJ01Rejection of invention patent application after publication

Application publication date:20140430

RJ01Rejection of invention patent application after publication

[8]ページ先頭

©2009-2025 Movatter.jp