CN103346944B - The looped network of multi-point cooperative detection link - Google Patents
The looped network of multi-point cooperative detection link Download PDFInfo
- Publication number
- CN103346944B CN103346944B CN201310241098.5A CN201310241098A CN103346944B CN 103346944 B CN103346944 B CN 103346944B CN 201310241098 A CN201310241098 A CN 201310241098A CN 103346944 B CN103346944 B CN 103346944B
- Authority
- CN
- China
- Prior art keywords
- node
- message
- looped network
- ring
- vice
- 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.)
- Active
Links
Abstract
The present invention relates to multi-point cooperative detection, multichannel looped network, node apparatus and detection method, wherein, from node, at least one is the vice-node having origination function concurrently；Host node and vice-node cooperation detection Link State, originate with certain time intervening sequences each other or originate looped network detection messages simultaneously, and the message that each node originates order in ring is forwarded, until returning the node originating this message, the CPU of described node apparatus and asic chip are constituted according to the type of message of transmission multiple passage, and various type of messages use the dedicated transmission channel of its correspondence.The technology of the present invention effect includes: significantly improve looped network resume speed and reliability, it is greatly saved ring node cpu calculation resources, and avoids packet loss phenomenon, when not increasing looped network node computing load, looped network scale is more big, and when looped network breaks down, the communication recovery time is more short.
Description
Technical field
The application relates to a kind of multi-point cooperative detection, multichannel looped network, node apparatus and detection method.
Background technology
Ring network technologies has important function for improving industrial network reliability, and looped network international standard has multiple, including RFC3619, IEC62439, ITU-G.8032 etc..At present, each manufacturer's looped network implementation is widely different, most looped network schemes are responsible for looped network configuration by host node, the cycle sends looped network detection messages, the message be responsible for checking direct-connected port links state from node, assist each link point of forwarding to send, detection messages finally raps around to host node, thus completing a ring status detection.The shortcoming of common looped network scheme is that looped network resume speed is heavily dependent on host node operational capability and from node-node transmission disposal ability; it is Millisecond that usual ring node cpu E-Packets consuming time; and ring node cpu real time execution multitask system; uncertain factor is more; easily cause the loss that E-Packets; therefore, along with the increase of looped network node number, the resume speed of looped network and reliability all can be a greater impact；It addition, port direct connected link state can only be detected by what CPU E-Packeted from node, if having change of medium equipment or trunking between link, then link failure cannot be detected from node.
For the improvement of these problems, at present more existing progress, for instance, CN200710121620.0, the application scheme such as CN200710000925.6, wherein CN200710121620.0 adds the self-healing ability of looped network, but unrealized dynamic distributed computing；CN200710000925.6 achieves looped network distributed arithmetic, however it is necessary that module during introducing hardware high accuracy pair, adds system failure point；It is more long for recovery time that both the above scheme all there is problems of the disconnection of looped network node scale more big then looped network.
Summary of the invention
The invention aims to solve above-mentioned technical problem of the prior art.
The technical solution used in the present invention is:
A kind of looped network of multi-point cooperative detection link, looped network include a host node, at least one from node, host node can origination, can E-Packet from node；It is characterized in that, described from node, at least one is the vice-node having origination function concurrently；Described host node and vice-node cooperation detection Link State, each other with certain time interval, order originates looped network detection messages, and message order in ring that each node originates is forwarded, until returning the node originating this message.
A kind of looped network of multi-point cooperative detection link, looped network include a host node, at least one from node, host node can origination, can E-Packet from node；It is characterized in that, described from node, at least one is the vice-node having origination function concurrently；Described host node and vice-node cooperation detection Link State, originate looped network detection messages simultaneously with one another, and message order in ring that each node originates is forwarded, until returning the node originating this message.
Preferably, all is all the vice-node having origination function concurrently from node.
Preferably, described host node is responsible for sending ring configuration detection messages, receives ring configuration detection messages from node and fills in own node information, forwarding this message, it is achieved looped network state collection；
When looped network is normal, described host node sends forward table reset message and ring configuration notification packet, notices ring status information, and vice-node receives message and sets the looped network detection time according to message instruction；
Subsequently, described host node and described vice-node sequence interval send or send ring status detection messages simultaneously；
When link failure, described vice-node sends warning message to described host node；When link-recovery, send link-recovery message to described host node.
Preferably, looped network node includes CPU and asic chip.
Preferably, looped network detection messages is sent out by asic chip master or forwards.
Preferably, described host node can configure detection messages, looped network configuration notification packet and forward table reset message by main hair ring net, can lead and sends out and forward looped network detection messages, can receive and process chain circuit alarming message and link-recovery message.
Preferably, described vice-node can be led and sends out and forward looped network detection messages, chain circuit alarming message and link-recovery message, can forward looped network configuration detection messages, looped network configuration notification packet and forward table reset message.
A kind of method of multi-point cooperative detection looped network, described looped network include a host node, at least one from node, host node can origination, can E-Packet from node；It is characterized in that, described in making, from node, at least one becomes the vice-node having origination function concurrently；Described host node and vice-node cooperation detection Link State, originate looped network detection messages simultaneously with one another, and message order in ring that each node originates is forwarded, until returning the node originating this message.
A kind of method of multi-point cooperative detection looped network, described looped network include a host node, at least one from node, host node can origination, can E-Packet from node；It is characterized in that, described in making, from node, at least one becomes the vice-node having origination function concurrently；Described host node and vice-node cooperation detection Link State, each other with certain time interval, order originates looped network detection messages, and message order in ring that each node originates is forwarded, until returning the node originating this message.
Preferably, all vice-nodes all becoming from node and having origination function concurrently are made.
Preferably, described host node sends ring configuration detection messages, receives ring configuration detection messages from node and fills in own node information, forwarding this message, it is achieved looped network state collection；
When looped network is normal, described host node sends forward table reset message and ring configuration notification packet, notices ring status information, and vice-node receives message and sets the looped network detection time according to message instruction；
Subsequently, described host node and described vice-node sequence interval send or send ring status detection messages simultaneously；
When link failure, described vice-node sends warning message to described host node；When link-recovery, send link-recovery message to described host node.
Preferably, after described ring configuration detection messages enters link point, message will only by transmitted to CPU, CPU send from looped network another port after making respective handling again.
Preferably, after described forward table reset message and ring configuration notification packet enter link point, simultaneously transmit identical messages copy to two inner tracks, first path be on send ring node cpu, second path is to be forwarded the packet to looped network another port by asic chip.
Preferably, described ring status detection messages, adopt asic chip pass-through mode when entering link point, forward from looped network another port, ring node cpu is sent on not, when this message loops back to the origination node of this message along looped network, can by this messages transmitted to CPU and terminate this message transmissions.
Preferably, described chain circuit alarming message and link-recovery message adopt asic chip pass-through mode when entering vice-node, forward from looped network another port, send ring node cpu on not, when this message is along looped network transmission to host node, can by this messages transmitted to CPU and terminate this message transmissions.
Preferably, described interval is that the ring configuration detection messages sent out by described host node master and processed successively by each vice-node obtains, and message comprises parameter: looped network detection cycle T h, ring node ID Ri；When ring configuration message returns to host node, add up link point total quantity Rc=Ri, calculate each link point assay intervals time Tc=Th/Rc successively according to Th and Rc.
A kind of looped network node device for realizing cooperation detection link, including: CPU, asic chip, it is characterised in that described CPU can process and transmit ring configuration detection messages, chain circuit alarming message and link-recovery message；Described asic chip while CPU processing forward table reset message and ring configure notification packet, can forward forward table reset message and ring configuration notification packet；Described asic chip can be led and send out and forward ring status detection messages, and can lead send out, forwarding and process chain circuit alarming message and link-recovery message；
The CPU of described node apparatus constitutes different communication channels with asic chip according to the type of message of transmission, and these communication channels include cpu access, the passage being made up of parallel CPU and asic chip, asic chip passage.
A kind of multichannel looped network node device, including: CPU, asic chip, it is characterized in that, the CPU of described node apparatus constitutes different communication channels with asic chip according to the type of message of transmission, these communication channels include cpu access, the passage being made up of parallel CPU and asic chip, asic chip passage；Described asic chip passage is for transmission looped network detection messages, chain circuit alarming message and link-recovery message.
Preferably, can process and transmit ring configuration detection messages, chain circuit alarming message and link-recovery message including: described CPU；Described asic chip while CPU processing forward table reset message and ring configure notification packet, can forward forward table reset message and ring configuration notification packet；Described asic chip can be led and send out and forward ring status detection messages, and can lead send out, forwarding and process chain circuit alarming message and link-recovery message.
A kind of looped network node device for realizing cooperation detection link, including: CPU, asic chip, it is characterized in that, the looped network configuration message received can be updated its time and number of nodes parameter by the CPU of described node apparatus, can according to the looped network configuration notification packet received and forward table reset message, configuration sends the time parameter of message, empties forward table；Described node apparatus has the transmission control module controlling message in setting interval transmission, and has the module that the ring detection messages of reception carries out forward；Described node apparatus has two ports, and wherein the first port sends message, and the second port receives the message being sent from the first port by this node and returning around ring.
Preferably, described asic chip is for transmission looped network detection messages, chain circuit alarming message and link-recovery message.
Preferably, described node apparatus is vice-node device.
Preferably, described node apparatus is switch.
The technology of the present invention effect is: significantly improve resume speed and the reliability of looped network.
The asic chip time delay that E-Packets is very little, reaches Microsecond grade and postpones, thus looped network it doesn't matter with looped network node quantity size recovery time.And looped network node quantity is more many, detection interval Tc is more little for looped network, then looped network is more short for recovery time.
The present invention supports multiple passage pass-through mode, and the mode such as including CPU software forwarding and asic chip forwarding, asic chip forwards can complete detection process in hundreds of microseconds, is prevented effectively from the accidental packet loss problem that software forward mode occurs.
Without accurate timing agreement or module between the major-minor node of looped network of the present invention, by self-adapting synergizing algorithm, it may be achieved the cooperation detection adapted with network size, when not increasing ring node cpu computational burden, looped network scale can be realized more big, the effect that looped network recovery time is more short.
The present invention can realize the cooperation detection adapted with network size, when ring number of nodes is very huge, if employing dot interlace or the mode every multiple points detect looped network state, then can avoid excessively frequent looped network status monitoring,
Microsecond grade precise synchronization can be realized by asic chip pass-through mode, thus supporting the collaborative operation of looped network equipment, for instance equipment synchronizes to restart between the major-minor node of looped network of the present invention.
The major-minor node of looped network of the present invention, without detecting looped network port linkdown/linkup event, has been greatly saved ring node cpu calculation resources.Avoid the defect that cannot detect linkdown/linkup event when physical link exists relaying.
Accompanying drawing explanation
Fig. 1 is multi-point cooperative ring-network topology figure of the present invention.
Fig. 2 is the link point first passage schematic diagram for transmitting ring configuration detection messages.
Fig. 3 is the link point second channel schematic diagram for transmitting ring configuration notification packet, forward table reset message.
Fig. 4 is for transmitting the link of ring status detection messages point third channel schematic diagram.
Fig. 5 is the link point fourth lane schematic diagram for transmission link warning message, link-recovery message.
Fig. 6 is multi-point cooperative ring network host node flow chart.
Fig. 7 is multi-point cooperative looped network vice-node flow chart.
Detailed description of the invention
Below in conjunction with accompanying drawing, the specific embodiment of the invention is described, but the invention is not restricted to this.Present invention resides in the various schemes described in description and the reasonable combination by the various measures mentioned in this specification.The relevant term of the present invention is as follows:
Host node (Master): specially appointed exclusive node in a loop, is responsible for collecting each transmission node information in ring, issues configuration parameter and control instruction to vice-node.
Vice-node (Assistant): node except Master in looped network, is responsible for requiring to report status information the coordination instruction according to host node according to host node, in order and independent detection looped network state.Why called after vice-node be in order to be different from common looped network from node, in the present invention, vice-node has the function of part host node.
Master port (MainPort): host node working port, is responsible for data transmission when looped network is normal.
Standby port (backupPort): host node backup port, when network failure occurs in master port, will enable the port and carry out data transmission.
Ring configuration detection messages (RingcheckMessage): host node timing sends to looped network, in order to collect ring status information, vice-node receives message and fills in own node information, forwards this message.
Ring configuration notification packet (RingcheckMessage): host node timing sends to looped network, in order to notice ring status information, vice-node receives message and sets the looped network detection time according to message instruction.
Ring status detection messages (HelloMessage): host node and vice-node timing constantly send to looped network, in order to detection loop health status.
Chain circuit alarming message (LinkWarmMessage): after vice-node detects link failure, to the alarm that host node is initiated.
Link-recovery message (LinkupMessage): after vice-node detects link failure recovery, to the notice that host node is initiated.
Forward table reset message (ResetFDBMessage): after host node confirms link failure, the removing forward table initiated to vice-node is noticed.
VLAN: virtual subnet.
Multi-point cooperative looped network of the present invention includes a host node and multiple vice-node.Host node is specially appointed exclusive node in ring, is responsible for collecting all vice-node information in ring, issues configuration parameter and control instruction to vice-node, detects looped network state；Vice-node is responsible for performing the coordination instruction of host node, and each vice-node also is responsible for detection looped network state, if be detected that the change of looped network state, then sends out message notifying host node.Host node and vice-node all have transmission, forward ring status detection messages function, and host node, vice-node are such as through sending ring status detection messages independent detection looped network state.Host node and each vice-node in a certain order, interval Preset Time successively, send ring status detection messages.The detection messages that each link point (include host node, vice-node, lower with) sends is through each descending link point, until returning to itself.The present invention adopts adaptive algorithm (described below), the Preset Time making each link point transmission ring status detection messages is relevant with looped network node quantity size, and looped network scale is more big, and detection messages Preset Time is more little, thus adding the frequency of detection, find failed link more in time.
Along with the increase of detection messages transmission frequency, the forward efficiency of detection messages becomes outstanding problem, so the present invention proposes following solution for this: by dividing the way of multiple transmission channel, various type of messages use the dedicated transmission channel of its correspondence.Wherein, the type of message that altofrequency sends is forwarded by link point asic chip, and the message that low frequency sends is forwarded by link point CPU software.
The looped network message that link point sends enters next link point, the message that link point receives includes at least one in following six kinds of messages, these six kinds of messages are: ring configuration detection messages, forward table reset message, ring configuration notification packet, ring status detection messages, chain circuit alarming message, link-recovery message.These six kinds of messages are divided into four kinds of type of messages, and every kind of type of message, by selecting corresponding transmission channel, has four kinds of transmission channels, respectively first, second, third, fourth transmission channels.
The first type of message comprises ring configuration detection messages, as in figure 2 it is shown, after this type message entrance link point, message will only by transmitted to CPU, CPU send from looped network another port after making respective handling again.
The second type of message comprises forward table reset message and ring configuration notification packet, as shown in Figure 3, after this type message enters link point, transmit identical message copy to two inner tracks simultaneously, ring node cpu is sent on being in first path, and second path is to be forwarded the packet to looped network another port by asic chip.
The third type of message comprises ring status detection messages, as shown in Figure 4, asic chip pass-through mode is adopted when entering link point, forward from looped network another port, ring node cpu is sent on not, when this message loops back to the origination node of this message along looped network, can by this messages transmitted to CPU and terminate this message transmissions.
4th kind of type of message comprises chain circuit alarming message and link-recovery message, as shown in Figure 5, this type message adopts asic chip pass-through mode when entering vice-node, forward from looped network another port, ring node cpu is sent on not, when this message is along looped network transmission to host node, can by this messages transmitted to CPU and terminate this message transmissions.
The method dividing transmission channel according to type of message is advantageous in that:
Ring configuration detection messages message adopts the first transmission channel, each ring node cpu can be made can to receive and process message, each link point can obtain required relevant information, host node can collect looped network integrality information, simultaneously, because first passage message amount is less and occurs in configuration phase, so not affecting for loop-net operation stage performance.
The second message transmitting method is advantageous in that the asic chip time delay that E-Packets is very little, reach Microsecond grade, it is negligible, so that all looped network vice-nodes almost receive forward table reset message and ring configuration notification packet simultaneously, receive forward table reset message, then all vice-nodes can start to empty mac address table immediately, avoiding conventional looped network adopts each ring node separation Millisecond time that CPU pass-through mode causes to empty mac address table successively, make whole net empty mac address table consuming time longer, and can not synchronize to empty mac address table between each link point；Work characteristics according to Ethernet switch, all switches empty MAC Address simultaneously and are rapidly reached steady statue most beneficial for network, it is to avoid network oscillation；Receive ring configuration notification packet, it is possible to making all vice-nodes synchronize to be clocked into state, provide time base for multiple link point cooperation detection looped network states, this advantage will be used below.
The third message transmitting method is advantageous in that ring status detection messages message amount is many and frequent, the transmission speed of this type message directly determines looped network detection performance, adopt the asic chip time delay that E-Packets very little, reach Microsecond grade, it is negligible, so that looped network detection required time is unrelated with looped network node quantity size, 10 equipment looped networks and 100 equipment looped network detection times are almost identical.
Owing to chain circuit alarming message and link-recovery message are burst messages, this type message needs with prestissimo transmission to ring network host node, adopts the asic chip time delay that E-Packets very little, reaches Microsecond grade, ring network host node it is negligible, so that can immediately receive Link State Packet.
The feature of comprehensive above four kinds of looped network type of messages and corresponding transmission channel, known ring status detection messages, chain circuit alarming message, link-recovery message can postpone to be transmitted with Microsecond grade, these three message transmissions can be approximately considered and postpone unrelated with looped network node quantity size, thus, the recovery time of looped network is unrelated with looped network node quantity size.
The host node of multi-point cooperative looped network of the present invention, vice-node detect looped network state jointly, and host node sends ring configuration detection messages at initial phase.Message comprises parameter: looped network detection cycle T h, ring node ID Ri.Wherein, Ri initial value=1.Vice-node receives this message, obtains looped network detection cycle T h, ring node ID Ri, then updates this message parameter, makes Ri=Ri+1, and this message is sent from another looped network port.Each vice-node processes ring configuration detection messages successively, and last vice-node forwards the packet to host node after having processed this message.
Host node receives ring configuration detection messages (comprising parameter Th, Ri) that each vice-node processed successively, adds up link point total quantity Rc=Ri, calculates each link point assay intervals time Tc=Th/Rc successively according to Th and Rc.Host node sends ring configuration notification packet (comprising parameter Th, Rc, Tc), and ring configuration notification packet is propagated by the second transmission channel, and each ring node cpu of this message up sending is forwarded to looped network another port by asic chip simultaneously.Advantage according to aforementioned second transmission channel, all vice-nodes almost receive ring configuration notification packet in the same time.
After each vice-node receives ring configuration notification packet, according to own node sequence number Ri and Tc, ring state-detection message sends initial time Ts=Ri*Tc, immediately begin to timing, after waiting the Ts time, sending ring status detection messages, after message sends, this vice-node repeats to send ring status detection messages every Th.
Ring status detection messages is by the 3rd transmission channel transmission; namely asic chip pass-through mode is adopted to propagate around ring; after traversal ring node, this message raps around to origination node; and transmitted to CPU; owing to adopting asic chip mode; under ring normal condition, this message almost can be back to origination node after sending immediately, does not have usual looped network and adopts CPU to forward the existing accidental packet loss problem of software.
Each link point basic synchronization receives ring configuration notification packet and starts timing, with the Tc time for interval between each link point, sends ring status detection messages successively, is used for detecting the whole net state of looped network.
Assume that a certain link of looped network occurs abnormal, namely have a certain link point to send ring status detection messages after 0 ~ Tc time, can detect that Link State is abnormal, send chain circuit alarming message immediately to host node.If being detected by host node self abnormal, not needing to send message, but warning count value adds 1.Chain circuit alarming message, by the 4th transmission channel transmission, adopts asic chip pass-through mode, receives destination address and is appointed as host node.
For avoiding the by mistake alarm of single link point, host node receive two (containing) after above chain circuit alarming message, standby port is set to forwarding state from blocked state；It is simultaneously emitted by forward table reset message (comprising ring status parameter RingFail).
Forward table reset message is by the 4th transmission channel transmission, adopt asic chip pass-through mode, all ring node cpus are sent on simultaneously, vice-node receives forward table reset message and then empties mac address table immediately, and link point is placed in RingFail state, still send ring status detection messages with Th for the cycle, if origination node is less than receiving along looped network wraparound message at the appointed time such as 1ms, then judging that looped network is still in abnormality, link point does not do any process.If receiving the ring status detection messages of wraparound in origination node at the appointed time such as 1ms, then judging that looped network is switched to normal condition from abnormality, sending link-recovery message immediately.
For avoiding the by mistake alarm of single link point, host node receive two (containing) after above link-recovery message, standby port is set to blocked state from forwarding state；It is simultaneously emitted by forward table reset message, this message (comprise ring status parameter RingOK) and adopt asic chip pass-through mode and on send all by way of vice-node CPU, vice-node receives forward table reset message and then empties mac address table, and link point is placed in RingOK state, still send ring status detection messages with Th for the cycle, detect that changing abnormal state then sends chain circuit alarming message.
Usually, two key indexs affecting looped network restorability are: consuming time needed for the frequency of ring status detection, the inspection of single ring status.Owing to transmitting ring configuration notification packet by the second transmission channel, each vice-node almost synchronizes to receive this message (having delicate level error), therefore, without accurate timing agreement or module between each vice-node, by self-adapting synergizing algorithm Tc=Th/Rc, can realizing looped network node scale more big, between link point, assay intervals Tc is more little, and namely ring status detection frequency is more high.Owing to transmitting ring status detection messages by the 3rd transmission channel so that each link point can complete the state-detection to whole ring within the Microsecond grade time, thus single ring detection process is unrelated with looped network node scale.To sum up, under the condition of similarly configuring, looped network scale is more big, and the frequency of ring status detection is more high, single ring status inspection the same (Microsecond grade difference can be ignored) required consuming time, thus, looped network is more short for recovery time, as shown in Figure 1.
The present invention preferentially adopts the multiple transmission channel of VLAN model split.Functional characteristics according to VLAN is it can be seen that the message come in from arbitrary port belonging to VLAN, if the match is successful for MAC Address, then this message is forwarded to port specified by MAC Address, if it fails to match for MAC Address, then this message will be broadcasted from port belonging to VLAN.The present embodiment will utilize functionality of vlan to realize four looped network transmission channels, create two and controls VLAN, number respectively VLAN1, VLAN2, create a service VLAN, be numbered VLAN3.VLAN1 has two looped network ports and CPU place port under its command, and VLAN1 will be used for first passage, second channel.VLAN2 has two looped network ports under its command, and VLAN2 will be used for third channel, fourth lane.VLAN3 is service VLAN, looped network user arrange according to practical situation.Host node, vice-node are adopted and are created two control VLAN and service VLAN in the same way, but host node and vice-node are slightly distinguished for the MAC Address set by each VLAN.The planning of the present embodiment MAC Address is as follows: all link points arrange port belonging to MAC-1, MAC-1 at VLAN1 and are appointed as CPU place port, and MAC-1 will be used for first passage；Host node arranges port belonging to MAC-2, MAC-2 at VLAN1 and is appointed as CPU place port, and MAC-2 is used for second channel；Vice-node is without arranging MAC-2；Links point arranges the proprietary MAC-3 address of each node at VLAN2, for instance host node is MAC-3A, and vice-node is followed successively by MAC-3B, MAC-3C etc., for ease of statement, MAC-3 address range is referred to as MAC-3X, port belonging to MAC-3X is appointed as CPU place port, and MAC-3X is used for third channel；Host node arranges port belonging to MAC-4, MAC-4 at VLAN2 and is appointed as CPU place port, and MAC-4 is used for fourth lane；Vice-node is without arranging MAC-4.
Work process of the present invention is as follows.
After host node starts, perform initialization operation, obtain configuration parameter, it is determined that ring detection messages sends cycle T h millisecond；Link dotted state is set to RingInit state.Create VLAN by preceding method and concerned MAC address is set.The looped network master port of VLAN1, VLAN2, standby port are set to forwarding state by host node, and the looped network standby port of VLAN3 is set to blocked state.Primary node status is in RingInit state.
After vice-node starts, perform initialization operation, obtain configuration parameter, create VLAN by preceding method and concerned MAC address is set.Two ports of looped network of VLAN1, VLAN2 are set to forwarding state by vice-node, and two ports of looped network of VLAN3 are set to forwarding state.Vice-node state is in RingInit state.
Host node sends ring configuration detection messages (comprising parameter: looped network detection cycle T h, ring node ID Ri=1) in VLAN1, and message MAC Address is MAC-1.
Vice-node receives ring configuration detection messages at VLAN1, obtains Th parameter, obtains this node ID Ri, then updates this message parameter Ri=Ri+1, and this message is sent from another looped network port.Each vice-node processes ring configuration detection messages successively, and last vice-node forwards the packet to host node after having processed this message.
Host node receives ring configuration detection messages (comprising parameter Th, Ri) that each vice-node processed successively, adds up link point total quantity Rc=Ri, calculates each link point assay intervals time Tc=Th/Rc successively according to Th and Rc.Link dotted state is set to RingOK state.Host node sends ring ring configuration notification packet (comprising parameter Th, Rc, Tc), and message address is MAC-2, and this message is directly propagated by asic chip pass-through mode in ring, and message send each vice-node CPU on simultaneously, and this message is along around being back to host node.
Each vice-node is after VLAN1 receives ring configuration notification packet, and immediately according to own node sequence number Ri and Tc, ring state-detection message sends initial time Ts=Ri*Rc, and link dotted state is set to RingOK state.Etc. after the Ts time sends ring status detection messages, repeat to send ring status detection messages every Th.
Ring status detection messages address is MAC_3X, and this message is propagated in ring by asic chip pass-through mode, and after traversal ring node, message raps around to origination node transmitted to CPU, terminates this message and propagates.Each link point detects looped network state with Tc successively for duration interval.
Assume that a certain link of looped network occurs abnormal, after 0 ~ Tc duration, namely can detect that link is abnormal, after any vice-node detects looped network abnormal state, send chain circuit alarming message immediately to host node, chain circuit alarming message adopts asic chip pass-through mode, and receiving destination address is MAC_4 address.After host node detects looped network abnormal state, update looped network detection anomaly statistics value Fc=Fc+1, do not send message.
Host node receives chain circuit alarming message, updates looped network detection anomaly statistics value Fc=Fc+1；For avoiding the alarm by mistake of single link point, Fc value more than or equal to 2, then judges that looped network is already at abnormality, and from blocked state, host node VLAN3 looped network standby port is set to forwarding state；Link dotted state is set to RingFail state.It is simultaneously emitted by forward table reset message (comprising ring status parameter RingFail)；It is 0 that host node Fc statistical value empties.
Forward table reset message (comprises ring status parameter RingFail), address is MAC-2, adopt asic chip pass-through mode, and on send all vice-node CPU, vice-node receives forward table reset message and then empties mac address table, and link point is placed in RingFail state, still sends ring status detection messages with Th for the cycle, detect that ring status continues extremely then not retransmit chain circuit alarming message, detect that ring status recovers normal and then sends link-recovery message.After host node detects that looped network recovering state is normal, update looped network detection normal statistics value Sc=Sc+1, do not send message.
Host node receives link-recovery message, updates looped network detection normal statistics value Sc=Sc+1；For avoiding the alarm by mistake of single link point, Sc value more than or equal to 2, then judges that looped network reverts to normal condition from abnormality, and from forwarding state, host node VLAN3 looped network standby port is set to blocked state, and it is 0 that Sc statistical value empties；It is simultaneously emitted by forward table reset message (comprising ring status parameter RingOK).This message adopts asic chip pass-through mode, and on send all vice-node CPU, vice-node receives forward table reset message and empties mac address table immediately, and link point is placed in RingOK state, vice-node still sends ring status detection messages with Th for the cycle, detects that changing abnormal state then sends chain circuit alarming message.
Alternatively, it is possible to adopting ACL(AccesControlList, access controls list) mode reaches to divide the purpose of transmission channel.
Those skilled in the art on the basis of present inventive concept, can also be variously changed.Such as but not limited to, only a part of node from node is set to the such vice-node of the present invention, i.e. make it have main sending out and forwarding capability.Be separated with between being wherein likely between these vice-nodes one or more do not have main send out but there is the common from node of forwarding capability.
These are only the preferred embodiments of the present invention, be not limited to the present invention；To those skilled in the art, the present invention can have various modifications and variations, all within present invention spirit and principle, any amendment of making, equivalent replacement, improvement etc., should be included within the scope of the invention.
Claims (4)
1. a looped network for multi-point cooperative detection link, looped network include a host node, at least one from node, host node can origination, can E-Packet from node；It is characterized in that, described from node, at least one is the vice-node having origination function concurrently；Described host node and vice-node cooperation detection Link State, originate ring status detection messages with certain time intervening sequences each other, or originate ring status detection messages simultaneously with one another；And the message that each node originates order in ring is forwarded, until returning the node originating this message；Specifically:
Described host node is responsible for sending ring configuration detection messages, receives ring configuration detection messages from node and fills in own node information, forwarding this message, it is achieved looped network state collection；
When looped network is normal, described host node sends forward table reset message and ring configuration notification packet, notices ring status information, and vice-node receives message and sets the looped network detection time according to message instruction；
Subsequently, described host node and described vice-node sequence interval send or send ring status detection messages simultaneously；
When link failure, described vice-node sends chain circuit alarming message to described host node；When link-recovery, send link-recovery message to described host node.
2. looped network according to claim 1, it is characterised in that all is all the vice-node having origination function concurrently from node.
3. looped network according to claim 1 and 2, it is characterised in that also there is at least one feature being selected from feature: looped network node includes CPU and asic chip；Ring status detection messages is sent out by asic chip master or forwards；Described host node can configure detection messages, ring configuration notification packet and forward table reset message by main hair ring, can lead and sends out and forward ring status detection messages, can receive and process chain circuit alarming message and link-recovery message；Described vice-node can be led and sends out and forward ring status detection messages, chain circuit alarming message and link-recovery message, can forward ring configuration detection messages, ring configuration notification packet and forward table reset message.
4. looped network according to claim 3, it is characterized in that, also there is at least one feature being selected from feature: after described ring configuration detection messages enters link point, message will only by transmitted to CPU, CPU send from looped network another port after making respective handling again；After described forward table reset message and ring configuration notification packet enter link point, transmit identical message copy to two inner tracks simultaneously, ring node cpu is sent on being in first path, and second path is to be forwarded the packet to looped network another port by asic chip；Described ring status detection messages, adopt asic chip pass-through mode when entering link point, forward from looped network another port, on not, send ring node cpu, when this message loops back to the origination node of this message along looped network, can by this messages transmitted to CPU and terminate this message transmissions；Described chain circuit alarming message and link-recovery message adopt asic chip pass-through mode when entering vice-node, forward from looped network another port, ring node cpu is sent on not, when this message is along looped network transmission to host node, can by this messages transmitted to CPU and terminate this message transmissions；Described interval is that the ring configuration detection messages sent out by described host node master and processed successively by each vice-node obtains, and message comprises parameter: looped network detection cycle T h, ring node ID Ri；When ring configuration detection messages returns to host node, add up link point total quantity Rc=Ri, calculate each link point assay intervals time Tc=Th/Rc successively according to Th and Rc.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310241098.5A CN103346944B (en) | 2013-06-18 | 2013-06-18 | The looped network of multi-point cooperative detection link |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310241098.5A CN103346944B (en) | 2013-06-18 | 2013-06-18 | The looped network of multi-point cooperative detection link |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103346944A CN103346944A (en) | 2013-10-09 |
| CN103346944B true CN103346944B (en) | 2016-07-27 |
Family
ID=49281720
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310241098.5A Active CN103346944B (en) | 2013-06-18 | 2013-06-18 | The looped network of multi-point cooperative detection link |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103346944B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105790902B (en) * | 2014-12-22 | 2020-06-09 | 研祥智能科技股份有限公司 | Method and system for realizing redundant network card switching |
| CN104899173A (en) * | 2015-04-03 | 2015-09-09 | 深圳市前海安测信息技术有限公司 | Node network based on node connection chip and information transmission port determining method |
| CN105072009B (en) * | 2015-09-25 | 2018-08-24 | 北京博维亚讯技术有限公司 | A kind of hardware based MRP looped network redundancies processing system and processing method |
| CN105897507B (en) * | 2016-03-31 | 2019-09-17 | 杭州数梦工场科技有限公司 | The condition detection method and device of node device |
| US20180059648A1 (en) * | 2016-08-31 | 2018-03-01 | Esab Ab | Techniques for network charting and configuration in a welding or cutting system |
| CN107026691B (en) * | 2017-04-14 | 2020-02-11 | 云南电网有限责任公司电力科学研究院 | Method and system for automatically isolating power distribution optical fiber ring network fault |
| CN108737142B (en) * | 2017-04-21 | 2020-11-06 | 华为技术有限公司 | Method for adjusting node detection parameters, node and mesh network |
| CN107346604B (en) * | 2017-06-01 | 2019-04-16 | 重庆世纪之光科技实业有限公司 | Meter reading network and its intelligent electric meter and power fail warning monitoring method and device |
| CN111817940B (en) * | 2020-09-11 | 2021-01-05 | 之江实验室 | Industrial control ring network system and implementation method thereof |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101252426A (en) * | 2007-09-11 | 2008-08-27 | 北京东土科技股份有限公司 | Method for realizing high-reliability distributed redundant looped network |
| CN101989934A (en) * | 2009-08-06 | 2011-03-23 | 中兴通讯股份有限公司 | Method and system for data ring network fault detection and location |
| CN102158383A (en) * | 2010-02-11 | 2011-08-17 | 华为技术有限公司 | Data transmission method, device and system for E1 bidirectional ring network |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2009206891A (en) * | 2008-02-28 | 2009-09-10 | Nec Corp | Layer 2 ring network system and management method therefor |
-
2013
- 2013-06-18 CN CN201310241098.5A patent/CN103346944B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101252426A (en) * | 2007-09-11 | 2008-08-27 | 北京东土科技股份有限公司 | Method for realizing high-reliability distributed redundant looped network |
| CN101989934A (en) * | 2009-08-06 | 2011-03-23 | 中兴通讯股份有限公司 | Method and system for data ring network fault detection and location |
| CN102158383A (en) * | 2010-02-11 | 2011-08-17 | 华为技术有限公司 | Data transmission method, device and system for E1 bidirectional ring network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103346944A (en) | 2013-10-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103346944B (en) | The looped network of multi-point cooperative detection link | |
| Rodeheffer et al. | Automatic reconfiguration in Autonet | |
| JP5073812B2 (en) | Distributed Ethernet system and method for detecting faults based on the system | |
| CN101662421B (en) | Method and device for transmitting control message based on ethernet multi-ring network | |
| CN100454880C (en) | Method and system of realizing rig network protection | |
| CN100450036C (en) | Method and apparatus for preventing loop when RRPP and partial STP network damage recovery | |
| CN101958831B (en) | Ethernet ring network failure recovery method, Ethernet ring network and switching equipment | |
| EP3166246B1 (en) | Method to detect and to handle failures in the communication in a computer network | |
| CN103580886A (en) | Packet transport network protection switching device and method | |
| EP3029883B1 (en) | Network protection method and apparatus, next-ring node, and system | |
| CN102255757A (en) | Link switching method and device | |
| CN102130912B (en) | Rapid ring protection protocol (RRPP)-based multiple virtual local area network (VLAN) register protocol (MVRP) realization method and equipment | |
| CN101714939A (en) | Fault treatment method for Ethernet ring network host node and corresponding Ethernet ring network | |
| CN1825832B (en) | Ring rapid spanning tree protocol | |
| CN101989953A (en) | Method and equipment for sending bidirectional forwarding detection message | |
| US20090006650A1 (en) | Communication device, communication method, communication interface, and program product | |
| CN103227725A (en) | Method and device for dual-server backup of firewall | |
| WO2015168947A1 (en) | Path switching method and device | |
| CN105357095A (en) | Looped network link fault handling system and method based on bidirectional routing | |
| CN101436975A (en) | Method, apparatus and system for implementing rapid convergence in looped network | |
| CN102918809B (en) | The method of network and expanding element and operational network | |
| EP2858302B1 (en) | Connectivity check method of service stream link, related apparatus and system | |
| CN102932183A (en) | Double-up link failure processing method and device | |
| CN106656311A (en) | Data link backup apparatus and method | |
| CN105897580A (en) | Protocol-independent forwarding network rapid fault detection and flow protection switching method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |