(19)
(11)EP 2 640 012 A1

(12)EUROPEAN PATENT APPLICATION
published in accordance with Art. 153(4) EPC

(43)Date of publication:
18.09.2013 Bulletin 2013/38

(21)Application number: 11840296.5

(22)Date of filing:  18.05.2011
(51)Int. Cl.: 
H04L 12/437  (2006.01)
(86)International application number:
PCT/CN2011/074279
(87)International publication number:
WO 2012/062097 (18.05.2012 Gazette  2012/20)
(84)Designated Contracting States:
AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

(30)Priority: 09.11.2010 CN 201010538404

(71)Applicant: ZTE Corporation
Shenzhen, Guangdong 518057 (CN)

(72)Inventor:
  • ZHOU, Ji
    Shenzhen Guangdong 518057 (CN)

(74)Representative: Conti, Marco et al
Bugnion S.p.A. Via di Corticella, 87
40128 Bologna
40128 Bologna (IT)

  


(54)MULTI-RING ETHERNET NETWORK AND PROTECTION METHOD THEREOF


(57) The present invention provides a multi-ring Ethernet protection method and a multi-ring Ethernet, and the multi-ring Ethernet protection method includes the following steps. A master node in a master ring determines connectivity condition of the master ring and a state of a shared link (S502); the master node in the master ring notifies the information of the determined connectivity condition of the master ring and the determined state of the shared link to a master node in a slave ring (S504); the master node in the slave ring judges whether to open a slave port according to the notification information from the master node in the master ring (S506). The present invention enables the master node in the slave ring to open the blocked port conditionally to ensure that the link, which is physically connected between a node in the master ring and a node in the slave ring, is logically connected for the service traffic, and can solve the problem that the link, which is physically connected between a node in the master ring and a node in the slave ring, is logically disconnected for the service traffic.




Description

Technical Field



[0001] The present invention relates to the field of data communication, in particular to a multi-ring Ethernet protection method and a multi-ring Ethernet.

Background



[0002] With the rapid development of Ethernet technology, the application of Ethernet becomes wider and wider, and the protection switching function of Ethernet becomes extremely important. Sine conventional Spanning Tree Protocol (STP) technology of the ring protection cannot meet the requirement of rapid convergence during link switching, Request for Comments (RFC) 3619 defines an Ethernet Automatic Protection Switching (EAPS) method, through which convergence time can be controlled within 50ms.

[0003] The EAPS technology has been realized by many device manufacturers, and many private technologies have been derived too; however, all of them have some defects.

[0004] ZTE Ethernet Smart Ring (ZESR) is the enhanced improvement and supplement over the EAPS, which inherits the advantage of rapid convergence of EAPS layer-2 network and meanwhile realizes the networking capability in a complex ring network environment. In order to prevent a link shared by master and slave rings from impacting simultaneously the calculation of the master and slave rings, in the ZESR the calculation of the slave ring is performed based on a hypothesis that higher-layer ring is connected permanently, that is to say, that the shared link of the master and slave rings belongs to the master ring only, and the slave ring does no need to maintain the shared link. As shown in Fig. 1, the link between the Port 3 of S2 and the Port 3 of S3 belongs to the master ring, the master node of the ZESR master ring is S1, and the master node of the slave ring is S4. Normally, the master ring blocks the Port 2 of S1, the slave ring blocks the Port 1 of S4, communication can be performed among S1, S2, S3 and S4, and no service loop would appear. However, this hypothesis has a defect, that is, when the master ring is no longer connected, the slave ring would not open the blocked logical port, consequently, a node in the slave ring and a node in the master ring are connected in physical link but logically disconnected for the service traffic, thus, service traffic is not connected among them. As shown in Fig. 2, when the link between the Port 1 of S1 and the Port 2 of S2 has a failure and the link between the Port 3 of S2 and the Port 3 of S3 has a failure too, the Port 2 of S1 in the master ring is no longer blocked; however, the Port1 of S4 in the slave ring still keeps blocked, in this way, S1, S3 and S2, S4 becomes two independent logic areas, between which traffic is not connected.

[0005] However, Rapid Ring Protection Protocol (RRPP) of the ring protection technology considers the above problem; in this technology, the slave ring sends a detection message periodically; when the master ring is no longer connected, the slave ring would open the blocked port in the slave ring due to the timeout of the detection message. Besides, the RRPP also considers the condition of dual-home link, as shown in Fig. 3, a slave ring 1 blocks the Port 1 of S4 and a slave ring 2 blocks the Port 2 of S5. If the master ring is disconnected, the blocked ports of the slave ring 1 and the slave ring 2 are opened simultaneously; then, net storm is formed between the two slave rings. Therefore, as shown in Fig. 4, the RRPP would block an edge port of an edge node in the slave ring before opening the slave port of the master node in the slave ring, so as to avoid data loop between multiple slave rings. However, this method cannot essentially solve the problem that the physical link between a node in the master ring and a node in the slave ring is connected but the link between the two nodes is logically disconnected for the service traffic when the master ring itself is disconnected. Because, according to the implementation of the RRPP, when the master ring link is disconnected, the slave ring still has one port blocked; as shown in Fig. 2, when the link between S1 and S2 and the link between S2 and S3 have a failure, the RRPP would open the Port 1 of S4 and blocks the Port 1 of S2; although S1 and S4 can perform normal communication at this moment, S2 becomes an isolated node, and the communication between S2 and any one of S1, S3, and S4 would be interrupted. Further, if the slave port of the master ring becomes physically down from logic blocked, and the shared link of the master and slave rings is disconnected too, the Port 1 of S2 still is blocked in Fig. 1 according to the method of RRPP; in this way, S1 and S4 cannot communicate with each other. Therefore, the RRPP cannot well resolve the problem of disconnection of master ring in these abnormal conditions.

[0006] To sum up, in the above conventional art each area knows the link state of its own ring but does not know the link state of other rings, thus, the slave ring has no response when the link of the master ring itself is disconnected and consequently the link, which is physically connected between a node in the master ring and a node in the slave ring, is logically disconnect for the service traffic, therefore, the problem, that the physical link between a node in the master ring and a node in the slave ring is connected but the link between the two nodes is logically disconnected for the service traffic when the master ring itself is no longer connected, cannot be well solved.

Summary



[0007] The present invention is to provide a multi-ring Ethernet protection method and a multi-ring Ethernet to at least solve the above problem.

[0008] According to one aspect of the present invention, a multi-ring Ethernet protection method is provided, which includes: determining, by a master node in a master ring, connectivity condition of the master ring and a state of a shared link; notifying, by the master node in the master ring, the information of the determined connectivity condition of the master ring and the determined state of the shared link to a master node in a slave ring; judging, by the master node in the slave ring, whether to open a slave port according to the information.

[0009] Notifying the information of the connectivity condition of the master ring and the state of the shared link to the master node in the slave ring includes: the master node in the master ring sending a flush protocol message containing identification field to the master node in the slave ring, if the master node in the master ring determines that the master ring is disconnected and the state of the shared link is down, wherein the flush protocol message containing identification field carries one or more identification fields used for indicating the shared link and the number of identification fields; and judging, by the master node in the slave ring, whether to open the slave port according to the information includes: the master node in the slave ring receiving the flush protocol message containing identification field, determining that the master ring is disconnected when judging that the flush protocol message containing identification field carries the number of identification fields, and opening the slave port.

[0010] Notifying, by the master node in the master ring, the information of the determined connectivity condition of the master ring and the determined state of the shared link to the master node in the slave ring includes: the master node in the master ring sending a flush protocol message containing identification field to the master node in the slave ring, when determining that the master ring is connected and the state of the shared link is down, wherein the flush protocol message containing identification field carries one or more identification fields used for indicating the shared link and a first predetermined value used for indicating connectivity of the master ring; and judging, by the master node in the slave ring, whether to open the slave port according to the information includes: the master node in the slave ring receiving the flush protocol message containing identification field; the master node in the slave ring determining that the master ring is connected according to the first predetermined value carried in the flush protocol message containing identification field, and keeping blocking the slave port after determining that there is no link failure in the slave ring.

[0011] Determining, by the master node in the master ring, the connectivity condition of the master ring and the state of the shared link includes: after detecting a link-down of the shared link, a first boundary node sending a link-down protocol message containing no identification field to the master ring and simultaneously sending a link-down protocol message containing identification field to the slave ring, wherein the link-down protocol message containing identification field carries one or more identification fields and the number of identification fields, the carried identification field is a first identification field consisting of a Media Access Control (MAC) address of the first boundary node and an Identifier (ID) of a port sending out the link-down protocol message containing identification field; after a second boundary node at the other end of the first boundary node receives the link-down protocol message containing identification field, the second boundary node adding a second identification field in the link-down protocol message containing identification field and sending the link-down protocol message containing identification field to the master ring, wherein the second identification field includes an MAC address of the second boundary node and an ID of an ingress port of the link-down protocol message containing identification field; after receiving the link-down protocol message containing identification field, the master node in the master ring saving the first identification field and the second identification field to a local buffer area; the master node in the master ring judging whether the link-down protocol message containing no identification field and the link-down protocol message containing identification field are received on different ports respectively, according to locally recorded source MAC addresses and ingress port IDs of the received link-down protocol message containing no identification field and the received link-down protocol message containing identification field; if the link-down protocol message containing no identification field and the link-down protocol message containing identification field are received on different ports respectively, the master node in the master ring judging whether the source MAC address of the received link-down protocol message containing no identification field is matched with the MAC address in the first identification field in the local buffer area or the MAC address in the second identification field in the local buffer area; and if the source MAC address of the received link-down protocol message containing no identification field is matched with the MAC address in the first identification field in the local buffer area or the MAC address in the second identification field in the local buffer area, the master node in the master ring determining that the master ring is connected and the shared link is down; otherwise, the master node in the master ring determining that the master ring is disconnected and the shared link is down.

[0012] After the first boundary node sends the link-down protocol message containing identification field to the slave ring, the method further includes: after receiving the link-down protocol message containing identification field, the master node in the slave ring recording the first identification field and passes through the link-down protocol message containing identification field to the second boundary node; and judging, by the master node in the slave ring, whether to open the slave port according to the information includes: the master node in the slave ring receiving the flush protocol message containing identification field, determining that the master ring is disconnected when judging that the flush protocol message containing identification field carries the number of identification fields, and opening the slave port when determining that an MAC address in one identification field carried in the flush protocol message containing identification field is matched with an MAC address in one identification field locally recorded by the master node in the slave ring.

[0013] The method further includes: the first boundary node stopping sending the link-down protocol message containing identification field to the slave ring, when detecting that the shared link is recovered; the master node in the slave ring blocking the slave port, when not receiving the link-down protocol message containing identification field within a predetermined time and determining that there is no link failure in the slave ring.

[0014] The one or more identification fields carried in the flush protocol message containing identification field which is sent by the master node in the master ring are all identification fields in a local buffer area of the master node in the master ring.

[0015] Notifying, by the master node in the master ring, the information of the determined connectivity condition of the master ring and the determined state of the shared link to the master node in the slave ring includes: if the master node in the master ring determines that the state of the shared link still is down and other links in the master ring are all recovered, the master node in the master ring sending a flush protocol message containing identification field to the master node in the slave ring, wherein the flush protocol message containing identification field carries one or more identification fields used for indicating the shared link and a second predetermined value used for indicating connectivity of the master ring; and judging, by the master node in the slave ring, whether to open the slave port according to the information includes: the master node in the slave ring receiving a flush protocol message containing identification field; the master node in the slave ring determining that the master ring is connected when judging that the flush protocol message containing identification field carries the second predetermined value, and blocking the slave port when determining that there is no link failure in the slave ring.

[0016] According to another aspect of the present invention, a multi-ring Ethernet is provided, which includes: a master node in a master ring, configured to determine connectivity condition of the master ring and a state of a shared link, and to notify the information of the determined connectivity condition of the master ring and the determined state of the shared link to a master node in a slave ring; and the master node in the slave ring, configured to judge whether to open a slave port according to the information.

[0017] The master node in the master ring includes: a sending module, configured to send a flush protocol message containing identification field to the master node in the slave ring, when the master node in the master ring determines that the master ring is disconnected and the state of the shared link is down, wherein the flush protocol message containing identification field carries one or more identification fields used for indicating the shared link and the number of identification fields; and the master node in the slave ring includes: a receiving module, configured to receive the flush protocol message containing identification field from the master node in the master ring; a judgment module, configured to judge whether the flush protocol message containing identification field carries the number of identification fields, and, if the flush protocol message containing identification field carries the number of identification fields, to determine that the master ring is disconnected; and a control module, configured to open the slave port of the master node in the slave ring in the condition that the judgment module determines that the master ring is disconnected.

[0018] With the present invention, the master node in the master ring collects the state information of the links in the master ring and notifies the information to the master node in the slave ring, and then the master node in the slave ring determines whether to block the slave port or open the slave port according to the notification information, thereby enabling the master node in the slave ring to open the blocked port conditionally to ensure that the physical link between a node in the master ring and a node in the slave ring is connected and the link between the two nodes is logically connected for the service traffic, and well solving the problem that the link, which is physically connected between a node in the master ring and a node in the slave ring, is logically disconnected for the service traffic when the master ring itself is no longer connected.

Brief Description of the Drawings



[0019] Drawings, provided for further understanding of the present invention and forming a part of the specification, are used to explain the present invention together with embodiments of the present invention rather than to limit the present invention, wherein:

Fig. 1 shows a topology of a master ring-slave ring link with no failure according to relevant art;

Fig. 2 shows a topology of a master ring-slave ring link in which the master ring has a failure according to relevant art;

Fig. 3 shows a topology of a dual-home link with no failure according to relevant art;

Fig. 4 shows a topology of a dual-home link in which the master ring has a failure and the slave ring opens blocked ports according to relevant art;

Fig. 5 shows a flowchart of a multi-ring Ethernet protection method according to an embodiment of the present invention;

Fig. 6 shows an application topology of a multi-ring Ethernet protection method according to the first preferred embodiment of the present invention;

Fig. 7 shows an application topology of a multi-ring Ethernet protection method according to the second preferred embodiment of the present invention;

Fig. 8 shows an application topology of a multi-ring Ethernet protection method according to the third preferred embodiment of the present invention; and

Fig. 9 shows a structure diagram of a multi-ring Ethernet according to an embodiment of the present invention.


Detailed Description of Embodiments



[0020] The embodiments are described in conjunction with the drawings as follows. The embodiments of the present application and the features of the embodiments can be combined with each other if there is no conflict.

[0021] First, a brief introduction is provided for some terminologies referred in the following embodiments of the present invention. When multiples rings are intersected, one ring is called a master ring and other rings intersected the master ring are called slave rings; the node where the master ring and the slave ring are intersected are called a boundary node; the link shared by the master ring and the slave ring is called a shared link. The port on the boundary node connected with the master node of the slave ring is called an edge port.

[0022] Fig. 5 shows a flowchart of a multi-ring Ethernet protection method according to an embodiment of the present invention, and the method includes the following steps.

[0023] Step 502, a master node in a master ring determines connectivity condition of the master ring and a state of a shared link.

[0024] For example, the master node may detect states of links in the master ring to determine whether the master ring is connected, when the shared link has a failure.

[0025] Step 504, the master node in the master ring notifies information of the determined connectivity condition of the master ring and the determined state of the shared link to a master node in a slave ring.

[0026] Step 506, the master node in the slave ring judges whether to open a slave port of the master node in the slave ring according to the information notified by the master node in the master ring.

[0027] In this embodiment, the master node in the master ring may collect state information of the links in the master ring and notifies the information to the master node in the slave ring, and then the master node in the slave ring determines whether to block the slave port or open the slave port according to the notification information, thereby enabling the master node in the slave ring to open the blocked port conditionally to ensure that the physical link between a node in the master ring and a node in the slave ring is connected and the link between the two nodes is logically connected for the service traffic, and well resolving the problem that the link, which is physically connected between a node in the master ring and a node in the slave ring, is logically disconnected for the service traffic when the master ring itself is no longer connected.

[0028] In actual implementation, in order to enable the master node in the master ring to accurately determine the connectivity condition of the master ring and the state of the shared link, an embodiment of the present invention utilizes some reserved fields in an EAPS frame of RFC3619. One EAPS frame totally has reserved words which are 36 bytes; then, a plurality of identification fields with a format of MAC address + PORTID (port ID) can be filled in a message of EAPS protocol, wherein the identification fields are used to indicate the shared link with failure (or link-down); each of the identification fields also identifies a boundary node having a failure in the link, and uniquely identify a slave ring. If the MAC address occupies 6 bytes and the PORTID occupies 2 bytes, then, 4 identification fields may be filled in the message at most. In addition, a counter is needed to maintain the number of the filled identification fields, and this counter occupies 1 byte. In this way, Step 502 might include the following steps to determine the connectivity condition of the master ring and the state of the shared link.

[0029] Step 1, after detecting the link-down of the shared link, a first boundary node (any one boundary node) sends a link-down protocol message containing no identification field to the master ring and simultaneously sends a link-down protocol message containing identification field to the slave ring, wherein the link-down protocol message containing identification field carries an identification field and the number of identification fields, the carried identification field is a first identification field consisting of an MAC address of the first boundary node and an ID of a port sending out the link-down protocol message containing identification field; obviously, the number of identification fields should be 1 at this moment.

[0030] In actual application, the number of identification fields can be represented by the value of the counter added in the message.

[0031] That is, each boundary node in the EAPS ring monitors the shared link of the master ring and the slave ring; if finding that the shared link is down (failed), besides sending to the master ring a common link-down protocol message (that is, the link-down protocol message containing no identification field; in the EAPS protocol, this type of message is sent out by a transmission node when the link has a failure, with a purpose of informing the master node that the ring has a failure), the boundary node also needs to send a link-down protocol message of the master ring(that is, a link-down protocol message containing identification field) to all edge ports on the nodes in order to send the link-down protocol message to the slave ring, wherein the link-down protocol message sent to the slave ring carries an identification field; the boundary node fills the identification field consisting of its own MAC address and the ID of the edge port sending out this message into the link-down protocol message. All nodes in the slave ring pass through this message.

[0032] In the above, if the failure does not occur on the shared link, the boundary node does not send the link-down protocol message containing identification field to the slave ring, but sends the link-down protocol message containing no identification field to the master ring only.

[0033] After receiving the link-down protocol message containing no identification field, the master node in the master ring would record the source MAC and the ingress port ID of this message.

[0034] Step 2, after a second boundary node at the other end of the first boundary node receives the link-down protocol message containing identification field, the second boundary node adds a second identification field in the link-down protocol message containing identification field and sends the link-down protocol message containing identification field to the master node in master ring, wherein the second identification field includes an MAC address of the second boundary node and an ID of an ingress port of the link-down protocol message containing identification field.

[0035] That is, when the link-down protocol message containing identification field sent by the first boundary node enters the second boundary node at the other end of the first boundary node through the slave ring, this message is reported to a Central Processing Unit (CPU) to be processed; then, the second boundary node fills into this message a new identification field (that is, the second identification field), including the MAC address of the second boundary node and the ID of an edge port from which this message enters, and sets the number of identification fields as 2; and then the second boundary node sends out this message to the master ring.

[0036] Step 3, after receiving the link-down protocol message containing identification field, the master node in the master ring saves the identification fields contained in this message (that is, the first identification field and the second identification field) to a local buffer area.

[0037] In actual application, after receiving the link-down protocol message containing identification field, the master node in the master ring also records the ingress port ID and the source MAC of this message.

[0038] Step 4, the master node in the master ring judges whether the link-down protocol message containing no identification field and the link-down protocol message containing identification field are received on different ports respectively, according to the locally recorded source MAC addresses and ingress port IDs of the received link-down protocol message containing no identification field and the received link-down protocol message containing identification field.

[0039] Since the master node in the master ring would record the source MAC and ingress port ID of each received message, the master node in the master ring may determine the connectivity condition of the master ring and decide whether to send the flush protocol message containing identification field (in the EAPS, the flush protocol message is sent out by the master node when the ring changes, with a purpose of synchronizing the states of one or more transmission nodes and notifying the one or more transmission nodes to flush MAC) or the flush protocol message containing no identification field (that is, common flush protocol message), according to the ingress port, source MAC of the message and the information of the identification fields in the local buffer area. If the master node in the master ring only receives the link-down protocol message containing identification field, the master node in the master ring caches all the identification fields in the message and records the ingress port ID and source MAC address of the message, and then sends the flush protocol message containing no identification field; if the master node in the master ring also receives the link-down protocol message containing no identification field from another port, the master node in the master ring determines to send the flush protocol message containing identification field. That is, only when the link-down protocol message containing no identification field and the link-down protocol message containing identification field are received on different ports respectively, the master node in the master ring would send the flush protocol message containing identification field.

[0040] Step 5, if the link-down protocol message containing no identification field and the link-down protocol message containing identification field are received on different ports respectively, the master node in the master ring judges whether the source MAC address of the received link-down protocol message containing no identification field is matched with the MAC address in the first identification field in the local buffer area or the MAC address in the second identification field in the local buffer area.

[0041] That is, after determining to send the link-down protocol message containing identification field, the master node in the master ring also needs to extract the source MAC address of the received message containing no identification field to match the MAC address in the identification field in the local buffer area, so as to judge whether there is one or more failures in the master ring.

[0042] Step 6, if the source MAC address of the received link-down protocol message containing no identification field is matched with the MAC address in the first identification field in the local buffer area or the MAC address in the second identification field in the local buffer area, it is indicated that there is only one link failure in the master ring currently, and the master node in the master ring determines that the master ring is connected and the shared link is down; if not matched, it is indicated that there are multiple link failures in the master ring currently, and the master node in the master ring determines that the master ring is disconnected and the shared link is down.

[0043] After determining the connectivity condition of the master ring and the state of the shared link through Step 1 to Step 6 included in Step 520, the master node in the master ring needs to notify the information of the connectivity condition of the master ring and the state of the shared link to the master node in the slave ring so that the master node in the slave ring makes a decision on whether to open the slave port; in this way, Step 504 might include the followings. If the master node in the master ring determines that the master ring is disconnected and the state of the shared link is down (that is, the matched condition in Step 6 included in Step 502), the master node in the master ring sends the flush protocol message containing identification field to the master node in the slave ring (the flush protocol message containing identification field is passed to the master node in the slave ring through the boundary node), wherein the flush protocol message containing identification field carries one or more identification fields used for indicating the shared link (that is, all identification fields in the local buffer area of the master node in the master ring) and the number of identification fields; then, Step 506 includes the following. After receiving the flush protocol message containing identification field, the master node in the slave ring determines that the master ring is disconnected when judging that the flush protocol message containing identification field carries the number of identification fields, and opens the slave port. In actual application, the number of identification fields carried in the flush protocol message containing identification field may be represented by the value of the counter, namely, setting the identification field counter carried in the sent flush protocol message containing identification field as the number of identification fields in this message.

[0044] In another condition, if the master node in the master ring determines that the master ring is connected and the state of shared link is down in Step 504 (that is, the unmatched condition in Step 6 included in Step 502), the master node in the master ring sends the flush protocol message containing identification field to the master node in the slave ring (the flush protocol message containing identification field is passed to the master node in the slave ring through the boundary node; that is, after the boundary node receives the flush protocol message, if the flush protocol message carries the one or more identification fields, the boundary node passes through the flush protocol message to the slave ring), wherein the flush protocol message containing identification field carries the one or more identification fields used for indicating the shared link (that is, all identification fields in the local buffer area of the master node in the master ring) and a first predetermined value used for indicating the connection of the master ring (this predetermined value may be a regulated big value, such as 64, 127); then, in Step 506, after receiving the flush protocol message containing identification field, the master node in the slave ring determines that the master ring is connected according to the first predetermined value carried in the flush protocol message containing identification field, and keeps blocking the slave port after determining that there is no link failure in the slave ring. In actual application, the first predetermined value also may be represented by the value of the counter in the message, namely, setting the identification field counter carried in the sent flush protocol message containing identification field as a big value (that is, the first predetermined value).

[0045] That is, after receiving the flush protocol message containing identification field, the master node in the slave ring analyzes the one or more identification fields in the message so as to decide whether to block or open the slave port; if the value of the identification field counter carried in this message is a very big value (that is, the first predetermined value), it is indicated that the master ring is connected; thus, it is needed to block the slave port of the master node in the slave ring (if there is no link failure in the slave ring right now); if the value of the identification field counter in this message is not the regulated big value (but is the number of the identification fields in this message), and the MAC in the identification field carried in this message is matched with any one MAC address recorded previously, the master node in the slave ring opens the slave port.

[0046] In addition, it also may be judged whether the master ring has one link failure only, according to whether the master node in the master ring receives from the master port and the slave port the link-down protocol message sent from one same MAC address respectively.

[0047] The preferred embodiments above describe the processing method performed when the link in the master ring has a failure.

[0048] During the process of actual implementation, after the master node in the master ring sends the link-down protocol message containing identification field to the slave ring in the step 1 above, the master node in the slave ring receives the link-down protocol message containing identification field, and needs to record the identification field (that is, the first identification field mentioned above) in this message and pass through the link-down protocol message containing identification field to the second boundary node; In order to avoid the occurrence of net storm; in the following Step 506, after receiving the flush protocol message containing identification field, the master node in the slave ring judges whether the flush protocol message containing identification field carries the number of identification fields in this message (the number can be represented by the value of the counter in the message); if the flush protocol message containing identification field carries the number of identification fields in this message, the master node in the slave ring determines that the master ring is disconnected; then the master node in the slave ring needs to judge whether the MAC address in the identification field carried in the flush protocol message containing identification field is matched with the MAC address of one identification field locally recorded by the master node in the slave ring; if so, the master node in the slave ring opens the slave port. In this way, the phenomenon, that a slave port of a master node in one slave ring is opened according to the flush protocol message containing identification field in other slave rings, is avoided when multiple slave rings coexist, thus net storm is prevented.

[0049] Obviously, since both boundary nodes (that is, the first boundary node and second boundary node) on two ends of the broken shared link would send the link-down protocol message containing identification field, and would add in the message the identification field consisting of the MAC address of itself and the ingress port ID of the message after receiving the link-down protocol message containing identification field sent from each other. At this moment, the master node in the slave ring may record the identification fields (that is, the first identification field and the second identification field) of two boundary nodes.

[0050] Finally, when the shared link is recovered, the boundary nodes on two ends of the shared link would stop sending the link-down protocol message containing identification field, if the master node in the slave ring does not receive the link-down protocol message containing identification field within a predetermined time (for example, link-down timeout time), and the slave ring itself does not have a link failure, the master node in the slave ring blocks the slave port. If the shared link still has a failure but other links in the master ring are recovered, the master node in the master ring would send the flush protocol message containing identification field (the one or more identification fields carried in the message may be all identification fields in the local buffer area of the master node in the master ring), wherein the value of the identification field counter in the flush protocol message containing identification field is a regulated big value (which is called a second predetermined value, for example, 64, 127 and the like), then the boundary node passes this message to the slave ring; after receiving this message, the master node in the slave ring analyzes that the value of the identification field counter in the message is the second predetermined value and then determines that the master ring is connected; if the slave ring itself has no link failure, the master node in the slave ring blocks the slave port. The above is the processing method performed when the link in the master ring is recovered.

[0051] Hereinafter, the specific processing method in the condition that the master ring is disconnected and the shared link is down, the master ring is connected and the shared link is down, or the shared link is recovered in the embodiment of the present invention is described according to the specific network topology shown in Fig. 6 to Fig. 8.

[0052] As shown in Fig. 6, in a master ring, a link between a Port 1 of a master node S1 and a Port 2 of a boundary node S2 has a failure (down), a link (shared link) between a Port 3 of the boundary node S2 and a Port 3 of a boundary node S3 has a failure too, thus the master ring is disconnected.

[0053] After the S1 knows the link failure at the Port 1, the S1 receives a link-down protocol message containing no identification field from itself immediately; at this moment, the S1 records that a source MAC of the link-down protocol message is MAC_S1 and an ingress port ID is 1; at this moment, the S1 opens an Port 2 and sends out a flush protocol message containing no identification field; after the S3 knows the link failure at the Port 3, the S3 sends the link-down protocol message containing no identification field to the Port 1 in the master ring immediately, wherein after the link-down protocol message enters the S1 from the Port 2, the S1 records that the source MAC of this message is MAC-S3 and the ingress port ID is 2; since one or more identification fields in the buffer area are empty, then the flush protocol message sent at this moment still contains no identification field; the S3 also finds that the Port 3 is in a shared link, thus, the S3 sends the link-down protocol message containing identification field to the Port 2, wherein the MAC is MAC_S3, the port ID is 2 and the counter is 1 in the identification field filled in this message; after the S4 receives this message, on one hand, the S4 sends out this message from the Port 2 intactly, on the other hand, the S4 records the identification field to its own protocol buffer area; after the S2 receives the message, the S2 fills its own identification field into this message, but the S2 finds that all master ring links where the S2 is located have a failure, thus this message is dropped; after the S2 finds the link failure at the Port 2 and Port 3, the S2 sends the link-down protocol message containing identification field to the Port 1, wherein the MAC is MAC_S2, the port ID is 1 and the counter is 1 in the identification field filled in the message; after the S4 receives this message, the S4 sends out this message from the Port 1 and records the identification field therein to its own protocol buffer area; after the S3 receives this message, the S3 fills its own identification field into this message, wherein the MAC is MAC_S3, the port ID is 2 and the counter becomes 2 in the new identification field; then, the S3 sends out this message from the Port 1; after the S1 receives this link-down protocol message, the S1 saves the identification fields carried in the message to its own buffer area and records that the source MAC is MAC_S2 and the ingress port ID is 2; when the S1 takes the source MAC in the link-down protocol message from the Port1 to match one or more identification fields in the buffer area, the S1 finds that match fails; therefore, the S1 sends the flush protocol message containing identification field at this time, wherein the identification field counter is a correct value, which is 2 at this moment. After the S3 receives the flush protocol message containing identification field, the S3 sends this message to the slave ring from the Port 2; after the S4 receives this message, the S4 compares identification fields previously saved in the protocol buffer area with the identification fields in the flush protocol message; then, MAC match succeeds and the value of the counter is 2; therefore, the S4 opens the slaved port 1.

[0054] As shown in Fig. 7, only a shared link of a master ring and a slave ring has a failure, that is, a link between a Port 3 of an S2 and a Port 3 of an S3 has a failure.

[0055] The S2 sends a link-down protocol message containing no identification field to an Port 2 and the S3 sends a link-down protocol message containing no identification field to an Port 1 too; after receiving these messages, the S1 sends the flush protocol message containing no identification field for one or more identification fields in the buffer area are empty. After detecting the failure of the shared link, both S2 and S3 send to the slave ring a master ring link-down protocol message containing respective identification field; and, after receiving the link-down protocol message from each other, both S2 and S3 fill respective identification field into the message and then send the message to the master ring; finally, the S1 can receive the link-down protocol messages containing identification field of both S2 and S3 from two ports; when the S1 takes source MACs of the link-down protocol messages from different ports to match with the identification fields, the S1 finds that they can be matched successfully and then sends the flush protocol message containing identification field, wherein a value of the identification field counter in the message is a big value (for example, 64), and both S2 and S3 forward the flush protocol message to the slave ring after receiving this message; after the S4 receives this message, the S4 parses out the value of the identification field counter in the message and finds that the highest position of this byte is 1, while a state of the slave ring is UP at this moment, thus, the S4 blocks the Port 1 and would not open it.

[0056] It should be noted that if the link state transforms into the state in Fig. 7 from the state in Fig. 6, that is, the shared link is always in down state but the link between S1 and S2 becomes recovered from disconnected, the Port 1 of the S1 and the Port 2 of the S2 need at least one link-down timeout+pre-up (pre-boot or pre-recover) time before entering forwarding state; during this time, the Port 1 of the S1 and the Port 2 of the S2 are temporarily blocked and the service traffic is not connected; therefore, this time is long enough for the S4 to receive the flush protocol message of which the value of the identification field counter is 64 to block the slave port; thus, net storm would not occur.

[0057] As shown in Fig. 8, only one non-shared link of a master ring has a failure, that is, a link between a Port 1 of an S1 and an Port 2 of an S2 has a failure.

[0058] At this moment, both S1 and S2 send a link-down protocol message containing no identification field in the master ring, and the S1 sends a flush protocol message containing no identification field only; therefore, these messages would not impact the slave ring.

[0059] It should be noted that if the link state transforms into the state in Fig. 8 from the state in Fig. 6, that is, the non-shared link of the master ring is always disconnected and the shared link between S2 and S3 is recovered, the Port 3 of the S2 and the Port 3 of the S3 need at least one link-down timeout+pre-up time before entering forwarding state; during this time, the Port 3 of the S2 and the Port 3 of the S3 are temporarily blocked and the service traffic is not connected. When the link is recovered, the Port 3 of the S2 and the Port 3 of the S3 stop sending all link-down protocol messages; if the master node S4 in the slave ring does not receive the link-down protocol message within a link-down timeout time, the S4 blocks the port 1; thus, net storm would not occur.

[0060] Fig. 9 shows a structure diagram of a multi-ring Ethernet according to an embodiment of the present invention, including: a master node 10 in a master ring, which is configured to determine connectivity condition of a master ring and a state of a shared link, and to notify the information of the determined connectivity condition of the master ring and the determined state of the shared link to a master node 20 in a slave ring; and the master node 20 in the slave ring, which is configured to judge whether to open a slave port according to the information.

[0061] During actual application, the master node 10 in the master ring may further include: a sending module 102, which is configured to send a flush protocol message containing identification field to the master node 20 in the slave ring, when the master node 10 in the master ring determines that the master ring is disconnected and the state of the shared link is down, wherein the flush protocol message containing identification field carries one or more identification fields used for indicating the shared link and the number of identification fields;
the master node 20 in the slave ring includes: a receiving module 202, which is configured to receive the flush protocol message containing identification field from the master node 10 in the master ring; a judgment module 204, which is configured to judge whether the flush protocol message containing identification field carries the number of identification fields, and, if so, to determine that the master ring is disconnected; a control module 206, which is configured to open the slave port of the master node 20 in the slave ring in the condition that the judgment module 204 determines that the master ring is disconnected.

[0062] In another condition, the sending module 102 of the master node 10 in the master ring is further configured to send a flush protocol message containing identification field to the master node 20 in the slave ring, when the master node 10 in the master ring determines that the master ring is connected and the state of the shared link is down, wherein the flush protocol message containing identification field carries one or more identification fields used for indicating the shared link and a first predetermined value (which may be a regulated big value, for example, 64. 127) used for indicating that the master ring is connected; and, the receiving module 202 of the master node 20 in the slave ring is further configured to receive the flush protocol message containing identification field from the master node 10 in the master ring; the judgment module 204 is further configured to judge whether the flush protocol message containing identification field carries the first predetermined value, and, if so, to determine that the master ring is connected, and to notify the control module 206 to keep blocking the slave port after determining that there is no link failure in the slave ring.

[0063] During actual application, in order to enable the master node to learn the connectivity condition of the master ring and the state of the shared link, the multi-ring Ethernet might include:

a first boundary node 30, which is configured to send a link-down protocol message containing no identification field to the master ring and simultaneously to send a link-down protocol message containing identification field to the slave ring, after detecting a link-down of the shared link, wherein the link-down protocol message containing identification field carries an identification field and the number of identification fields, the carried identification field is a first identification field consisting of an MAC address of the first boundary node and an ID of a port sending out the link-down protocol message containing identification field;

a second boundary node 40, which is located at the other end of the first boundary node 30 (the first boundary node 30 and the second boundary node 40 are connected through the shared link), and is configured to add, after receiving the link-down protocol message containing identification field from the first boundary node 30, a second identification field in the link-down protocol message containing identification field and to send the link-down protocol message containing identification field to the master ring, wherein the second identification field includes an MAC address of the second boundary node and an ID of an ingress port of the link-down protocol message containing identification field;

the master node 10 in the master ring is further configured to: save the first identification field and the second identification field to a local buffer area, after receiving the link-down protocol message containing identification field; and judge whether the link-down protocol message containing no identification field and the link-down protocol message containing identification field are received on different ports respectively, according to one or more locally recorded source MAC addresses and ingress port IDs of the received link-down protocol message containing no identification field and the received link-down protocol message containing identification field; if so, judge whether the MAC address in the first identification field or the MAC address in the second identification field in the local buffer area is matched with the one or more source MAC addresses of the received link-down protocol message containing no identification field; if so, determine that the master ring is connected and the shared link is down; otherwise, determine that the master ring is disconnected and the shared link is down.



[0064] At this moment, the one or more identification fields carried in the flush protocol message containing identification field which is sent by the sending module 102 of the master node 10 in the master ring are all identification fields in the local buffer area of the master node 10 in the master ring.

[0065] In the above, the master node 20 in the slave ring is further configured to record the first identification field after receiving the link-down protocol message containing identification field from the first boundary node 30, and to pass through the link-down protocol message containing identification field to the second boundary node 40; in this way, in order to avoid the occurrence of net storm, when judging whether to open the slave port of the master node 20, the master node 20 in the slave ring needs to judge that the received flush protocol message containing identification field carries the number of identification fields first and determines that the master ring is disconnected, and then determines that an MAC address in one identification field carried in the flush protocol message containing identification field is matched with an MAC address in one identification field recorded by the master node in the slave ring locally, and finally opens the slave port.

[0066] In the condition that the link in the master ring is recovered, the first boundary node 30 is further configured to stop sending the link-down protocol message containing identification field to the slave ring, when detecting that the shared link is recovered; then, the master node 20 in the slave ring blocks the slave port, when not receiving the link-down protocol message containing identification field within a predetermined time and when determining that there is no link failure in the slave ring.

[0067] In addition, in another link recovered condition that the shared link still is down but other links in the master ring are recovered, the sending module 102 of the master node 10 in the master ring is further configured to send the flush protocol message containing identification field to the master node 20 in the slave ring, when the master node 10 in the master ring determines that the state of the shared link still is down and other links in the master ring all are recovered, wherein the flush protocol message containing identification field carries the one or more identification fields (which may be all identification fields in the local buffer area of the master node in the master ring) used for indicating the shared link and a second predetermined value (which may be a regulated big value, for example, 64, 127) used for indicating the connection of the master ring; then, the judgment module 204 of the master node 20 in the slave ring is further configured to judge whether the flush protocol message containing identification field carries the second predetermined value after the receiving module 202 receives the flush protocol message containing identification field, and, if so, to determine that the master ring is connected, and then to notify the control module 206 to block the slave port when determining that there is no link failure in the slave ring.

[0068] From the description above, it can be seen that the present invention realizes the following technical effects: the protection and processing capability of the multi-ring Ethernet is enhanced through the improved EAPS; when the link of the master ring is no longer connected, the slave ring opens the blocked port conditionally to ensure that the link, which is physically connected between a node in the master ring and a node in the slave ring, is logically connected for the service traffic.

[0069] Obviously, those skilled in the art shall understand that the above-mentioned modules and steps of the present invention can be realized by using general purpose calculating device, can be integrated in one calculating device or distributed on a network which consists of a plurality of calculating devices. Alternatively, the modules and the steps of the present invention can be realized by using the executable program code of the calculating device. Consequently, they can be stored in the storing device and executed by the calculating device, or they are made into integrated circuit module respectively, or a plurality of modules or steps thereof are made into one integrated circuit module. In this way, the present invention is not restricted to any particular hardware and software combination.

[0070] The descriptions above are only the preferable embodiment of the present invention, which are not used to restrict the present invention. For those skilled in the art, the present invention may have various changes and variations. Any amendments, equivalent substitutions, improvements, etc. within the present invention are all included in the scope of the protection of the present invention.


Claims

1. A multi-ring Ethernet protection method, characterized by comprising:

determining, by a master node in a master ring, connectivity condition of the master ring and a state of a shared link;

notifying, by the master node in the master ring, information of the determined connectivity condition of the master ring and the determined state of the shared link to a master node in a slave ring; and

judging, by the master node in the slave ring, whether to open a slave port according to the information.


 
2. The method according to claim 1, characterized in that
notifying the information of the connectivity condition of the master ring and the state of the shared link to the master node in the slave ring comprises: the master node in the master ring sending a flush protocol message containing identification field to the master node in the slave ring, if the master node in the master ring determines that the master ring is disconnected and the state of the shared link is down, wherein the flush protocol message containing identification field carries one or more identification fields used for indicating the shared link and the number of identification fields; and
judging, by the master node in the slave ring, whether to open the slave port according to the information comprises: the master node in the slave ring receiving the flush protocol message containing identification field, determining that the master ring is disconnected when judging that the flush protocol message containing identification field carries the number of identification fields, and opening the slave port.
 
3. The method according to claim 1, characterized in that
notifying, by the master node in the master ring, the information of the determined connectivity condition of the master ring and the determined state of the shared link to the master node in the slave ring comprises: the master node in the master ring sending a flush protocol message containing identification field to the master node in the slave ring, when determining that the master ring is connected and the state of the shared link is down, wherein the flush protocol message containing identification field carries one or more identification fields used for indicating the shared link and a first predetermined value used for indicating connectivity of the master ring; and
judging, by the master node in the slave ring, whether to open the slave port according to the information comprises: the master node in the slave ring receiving the flush protocol message containing identification field; the master node in the slave ring determining that the master ring is connected according to the first predetermined value carried in the flush protocol message containing identification field, and keeping blocking the slave port after determining that there is no link failure in the slave ring.
 
4. The method according to any one of claims 1 to 3, characterized in that determining, by the master node in the master ring, the connectivity condition of the master ring and the state of the shared link comprises:

after detecting a link-down of the shared link, a first boundary node sending a link-down protocol message containing no identification field to the master ring and simultaneously sending a link-down protocol message containing identification field to the slave ring, wherein the link-down protocol message containing identification field carries one or more identification fields and the number of identification fields, the carried identification field is a first identification field consisting of a Media Access Control (MAC) address of the first boundary node and an Identifier (ID) of a port sending out the link-down protocol message containing identification field;

after a second boundary node at the other end of the first boundary node receives the link-down protocol message containing identification field, the second boundary node adding a second identification field in the link-down protocol message containing identification field and sending the link-down protocol message containing identification field to the master ring, wherein the second identification field comprises an MAC address of the second boundary node and an ID of an ingress port of the link-down protocol message containing identification field;

after receiving the link-down protocol message containing identification field, the master node in the master ring saving the first identification field and the second identification field to a local buffer area;

the master node in the master ring judging whether the link-down protocol message containing no identification field and the link-down protocol message containing identification field are received on different ports respectively, according to locally recorded source MAC addresses and ingress port IDs of the received link-down protocol message containing no identification field and the received link-down protocol message containing identification field;

if the link-down protocol message containing no identification field and the link-down protocol message containing identification field are received on different ports respectively, the master node in the master ring judging whether the source MAC address of the received link-down protocol message containing no identification field is matched with the MAC address in the first identification field in the local buffer area or the MAC address in the second identification field in the local buffer area; and

if the source MAC address of the received link-down protocol message containing no identification field is matched with the MAC address in the first identification field in the local buffer area or the MAC address in the second identification field in the local buffer area, the master node in the master ring determining that the master ring is connected and the shared link is down; otherwise, the master node in the master ring determining that the master ring is disconnected and the shared link is down.


 
5. The method according to claim 4, characterized in that
after the first boundary node sends the link-down protocol message containing identification field to the slave ring, the method further comprises: after receiving the link-down protocol message containing identification field, the master node in the slave ring recording the first identification field and passes through the link-down protocol message containing identification field to the second boundary node; and
judging, by the master node in the slave ring, whether to open the slave port according to the information comprises: the master node in the slave ring receiving the flush protocol message containing identification field, determining that the master ring is disconnected when judging that the flush protocol message containing identification field carries the number of identification fields, and opening the slave port when determining that an MAC address in one identification field carried in the flush protocol message containing identification field is matched with an MAC address in one identification field locally recorded by the master node in the slave ring.
 
6. The method according to claim 4, characterized by further comprising:

the first boundary node stopping sending the link-down protocol message containing identification field to the slave ring, when detecting that the shared link is recovered; and

the master node in the slave ring blocking the slave port, when not receiving the link-down protocol message containing identification field within a predetermined time and determining that there is no link failure in the slave ring.


 
7. The method according to claim 1 or 2, characterized in that the one or more identification fields carried in the flush protocol message containing identification field which is sent by the master node in the master ring are all identification fields in a local buffer area of the master node in the master ring.
 
8. The method according to claim 1, characterized in that
notifying, by the master node in the master ring, the information of the determined connectivity condition of the master ring and the determined state of the shared link to the master node in the slave ring comprises: if the master node in the master ring determines that the state of the shared link still is down and other links in the master ring are all recovered, the master node in the master ring sending a flush protocol message containing identification field to the master node in the slave ring, wherein the flush protocol message containing identification field carries one or more identification fields used for indicating the shared link and a second predetermined value used for indicating connectivity of the master ring; and
judging, by the master node in the slave ring, whether to open the slave port according to the information comprises: the master node in the slave ring receiving a flush protocol message containing identification field; the master node in the slave ring determining that the master ring is connected when judging that the flush protocol message containing identification field carries the second predetermined value, and blocking the slave port when determining that there is no link failure in the slave ring.
 
9. A multi-ring Ethernet, characterized by comprising:

a master node in a master ring, configured to determine connectivity condition of the master ring and a state of a shared link, and to notify information of the connectivity condition of the master ring and the determined state of the shared link to a master node in a slave ring; and

the master node in the slave ring, configured to judge whether to open a slave port according to the information.


 
10. The multi-ring Ethernet according to claim 9, characterized in that
the master node in the master ring comprises: a sending module, configured to send a flush protocol message containing identification field to the master node in the slave ring, when the master node in the master ring determines that the master ring is disconnected and the state of the shared link is down, wherein the flush protocol message containing identification field carries one or more identification fields used for indicating the shared link and the number of identification fields; and
the master node in the slave ring comprises: a receiving module, configured to receive the flush protocol message containing identification field from the master node in the master ring; a judgment module, configured to judge whether the flush protocol message containing identification field carries the number of identification fields, and, if the flush protocol message containing identification field carries the number of identification fields, to determine that the master ring is disconnected; and a control module, configured to open the slave port of the master node in the slave ring in the condition that the judgment module determines that the master ring is disconnected.
 




Drawing