WO2005060187A1 - クラスタシステム、クラスタメンバ、故障復旧方法及びプログラム - Google Patents
クラスタシステム、クラスタメンバ、故障復旧方法及びプログラム Download PDFInfo
- Publication number
- WO2005060187A1 WO2005060187A1 PCT/JP2004/018090 JP2004018090W WO2005060187A1 WO 2005060187 A1 WO2005060187 A1 WO 2005060187A1 JP 2004018090 W JP2004018090 W JP 2004018090W WO 2005060187 A1 WO2005060187 A1 WO 2005060187A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- session
- packet
- processing range
- sub
- processing
- Prior art date
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L69/00—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
- H04L69/40—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass for recovering from a failure of a protocol instance or entity, e.g. service redundancy protocols, protocol state redundancy or protocol service redirection
Definitions
- the present invention relates to a cluster system that functions as a router that forwards an IP bucket, and more particularly to a cluster system that can perform a failure recovery process without exchanging session information between cluster members.
- firewall devices used to block unauthorized access
- a VPN gateway device that terminates an IPsec tunnel.
- a cluster having the configuration shown in FIG. 20 can be considered as a router cluster for performing load distribution.
- the router cluster shown in the figure includes a plurality of router devices 101 to 10On having a session processing unit and a bucket sorting device 110 installed in front of them.
- the packet distribution device 110 realizes load distribution by distributing a packet input from the outside to any one of routers 100 to 10On in accordance with a predetermined load distribution rule.
- the router cluster shown in Fig. 20 has a problem that the load is concentrated on the packet distribution device 110, and if a failure occurs in the bucket distribution device 110, the entire system is paralyzed. There is a problem.
- a router cluster as shown in Fig. 21 has been proposed (for example, Table 2 00 3—5 1 7 2 2 1 and Table 2 0 0 3—5). 1 8 .3 See 38).
- the conventional router cluster shown in FIG. 21 includes one master router device 200 and a plurality of router devices (slave router devices) 201 to 200 n.
- Each of the routers 200 to 200 n includes a session processing unit and a traffic distribution filter.
- IP packets from the adjacent node 210 to the router cluster (sometimes simply referred to as buckets) are received by all routers 200 to 200n by multicasting in the data link layer protocol. .
- the traffic distribution filter in each device 200-0 to 20 ⁇ passes or discards the IP packet multicasted on the data link 220 according to the traffic distribution rule.
- the traffic distribution rules of the traffic distribution filter provided in each of the routers 200 to 20n satisfy the following conditions.
- the traffic distribution rules in the traffic distribution filter within each router 200 1 to 200 ⁇ are set by the master router 200, and the master router 200 It knows what traffic distribution rules are set in the traffic distribution filter within the router device 201 to 20 ⁇ , and the load of each router 210 to 200 ⁇ is equal. Set traffic distribution rules so that they are distributed.
- the master-router apparatus 200 has a traffic distribution filter for processing buckets that do not conform to the traffic distribution rules in its own apparatus. Further, the master router 200 sets the traffic distribution filter of the other routers 201 to 20 ⁇ that has generated a new traffic distribution rule from the session information of the processed packet. When a failure occurs in the master router 200, any one of the other routers 201 to 20 ⁇ operates as a master router.
- the session processing unit in each router device 200 to 20 ⁇ is set inside With reference to the specified session processing rules and session status, the bucket that has passed through the bucket distribution file is processed and discarded or transferred.
- the session processing rule of the session processing unit in each of the router devices 201 to 200n is set by the master device 200.
- the router devices 200 to 200 n including the master router device 200 mutually exchange session information indicating their own session state. This exchange of session information is performed at regular intervals, and each of the routers 200 to 200 n holds the session processing rule of the other router and the latest session state of the replaced router. . Therefore, when a failure occurs in any of the router devices 201 to 200 n, the master router device 200 determines an alternative device, and sets the processing rules and the processing rules set in the failed router device. It is possible to perform the process of passing the session state to the corresponding device. In addition, when a failure occurs in the master router 200, another router can take over the processing of the master router 200. Thus, according to the router class shown in Fig. 21, even if a failure occurs in any of the router devices 200 to 20n that constitute the router cluster, automatic recovery is possible. become.
- the conventional router class shown in Fig. 21 has the following problems. That is, since the session state of each router 201 to 200n is transmitted to the master router 200 at regular intervals, the other routers grasped by the master router 200 There may be a difference between the session state of 201 to 20n and the session state held by each router device 201 to 20n. Due to this difference, even if automatic failure recovery is performed by providing a substitute device, a packet that cannot be processed may be processed, or a packet that should be processed may not be processed. If the interval of exchanging session information is shortened to eliminate the above difference, the traffic for exchanging session information increases and the load of information exchange increases. Also, even if the number of router devices increases, the communication load increases. As a result, the conventional Lei-Cluster cluster could not simultaneously increase the number of routers and shorten the failover time, and there was a major problem in scalability. Disclosure of the invention
- an object of the present invention is to enable a failure recovery process without exchanging session information between class members.
- a first cluster system provides a class system comprising a plurality of cluster members
- a main processing range indicating a main processing range and a sub-processing range indicating a sub-processing range assigned to the own cluster member are determined in accordance with a predetermined range determination rule.
- a session process including a process of recording and updating session information indicating the state of the session to which the packet belongs and a bucket transfer process are performed, and the bucket that matches the sub-processing range determined above.
- the session processing including the record update processing of the session information indicating the state of the session to which the bucket belongs is performed, and the failure of the cluster member in charge of the main processing for the packet in which the own cluster member is in charge of the sub-processing has When a failure occurs, the cluster member in which the failure occurred was in charge of the main processing. It is characterized in that the main processing of the packet is taken over using the recorded session information.
- the second cluster system includes:
- a processing control unit Based on a predetermined range determination rule and the number of cluster members constituting the cluster system, distributed processing for determining a main processing range indicating a range of main processing and a sub-processing range indicating a range of sub-processing to be handled by a member of the own class.
- a traffic distribution filter that allows a packet that matches a main processing range or a sub-processing range determined by the distributed processing control unit to pass through among the packets input to the cluster system;
- a session information holding unit for storing session information indicating a state of a session to which the bucket belongs;
- a session process including a process of updating the session information of the packet stored in the session information holding unit is performed, and a bucket output process is performed.
- a bucket transfer unit for transferring the bucket output from the session processing unit
- a failure recovery unit that changes the sub-processing range to the main processing range when a failure occurs in the class component that is in charge of the main process for the bucket in which the own class component is in charge of the sub-process. It is characterized by.
- the third cluster system even when session information is not exchanged between class members, when a failed cluster member occurs or when a new cluster member is added, the main members in charge of each class member are added.
- the second class system In order to be able to change the processing range and sub-processing range, the second class system
- the distributed processing control unit When a failure occurs in another cluster member, and when a new cluster member is added to the cluster system, the distributed processing control unit performs a new main processing based on the range determination rule and the number of cluster members.
- the processing range and the sub-processing range are determined, and the main processing range and the sub-processing range up to that are set as the old main processing range and the old sub-processing range.
- the session processing unit may determine that the packet is a packet requesting to open a session, or the session information storage. On the condition that session information of the bucket is stored in the holding unit, a session process including an update process of session information of the bucket stored in the session information holding unit is executed, and a bucket output process is performed. Run the old lord If the packet matches only the processing range, the packet is discarded if the bucket is a packet requesting the opening of a session, otherwise, the session information of the bucket is stored in the session information holding unit.
- the bucket output processing is executed, and only the new sub-processing range is matched.
- the packet is stored in the session information holding unit on condition that the packet is a bucket requesting the opening of a session or that the session information of the packet is stored in the session information holding unit. Including a process of updating the session information of the bucket.
- the packet processing is executed together with the packet discarding processing, and if the bucket matches only the old sub-processing range, the packet is discarded if the bucket is a packet requesting to open a session; otherwise, the packet is discarded.
- session processing including updating the session information of the packet stored in the session information holding unit, on condition that the session information of the bucket is stored in the session information holding unit; For a bucket that performs bucket discarding processing and matches both the old sub-processing range and the sub-processing range, if the bucket is a bucket requesting to open a session, the session information holding unit Session processing including updating the session information of the packet stored in the And the bucket discarding process is executed, otherwise, the session information stored in the session information holding unit is stored on the condition that the session information of the bucket is stored in the session information holding unit. It is characterized in that it has a configuration for executing a session process including a process of updating session information of a bucket and executing a packet output process.
- the fourth cluster system according to the present invention is the third cluster system
- Each of the cluster members comprises a ballist
- An advertisement message including the identifier of the own cluster member is transmitted to other cluster members at predetermined time intervals, and when an advertisement message from another cluster member is received, the transmission source of the advertisement message is transmitted.
- An advertisement message processing unit that resets the life and death monitoring event corresponding to the cluster member,
- the distributed processing control unit controls the central processing control unit
- a class system characterized in that when a time-out alive monitoring timer occurs, a cluster member corresponding to the time-out alive monitoring timer is recognized as having failed.
- a fifth cluster system according to the present invention is the fourth cluster system, wherein the distributed processing control unit comprises:
- the identifier in the advertisement message received by the advertisement message processing unit is not included in the cluster member list, it is configured to recognize that the cluster member of the identifier has been added.
- the first cluster member according to the present invention performs failure recovery processing without exchanging session information between cluster members.
- the main processing range indicating the main processing range and the sub-processing range indicating the sub-processing range assigned to the own cluster member are determined according to a predetermined range determination rule.
- a session process including a record update process of session information indicating a state of a session to which the packet belongs and a bucket transfer process are performed.
- session processing including recording and updating of session information indicating the state of the session to which the packet belongs is performed.
- the first failure recovery method according to the present invention enables failure recovery processing to be performed without exchanging session information between cluster members.
- a main processing range indicating a main processing range and a sub-processing range indicating a sub-processing range assigned to the own cluster member are determined according to a predetermined range determination rule, and among the packets input to the cluster system, For a packet that matches the main processing range, session processing including recording and updating of session information indicating the state of the session to which the packet belongs and packet transfer processing are performed, and for packets that match the determined sub-processing range. Performs the session processing including the record update processing of the session information indicating the state of the session to which the packet belongs, and the failure of the cluster member in charge of the main processing for the bucket in which the own cluster member is in charge of the sub processing. When a failure occurs, the failed cluster member is in charge of the main processing. The main processing of the packet is succeeded by using the recorded session information.
- the first program according to the present invention is designed to perform a failure recovery process without exchanging session information between cluster members.
- a program for causing a computer to function as a cluster member, the computer comprising:
- a main processing range indicating the main processing range and a sub-processing range indicating the sub-processing range assigned to the own class member are determined according to a predetermined range determination rule, and input to the class system including the own cluster member as a component.
- the packet processing includes a session process including a record update process of session information indicating a state of a session to which the bucket belongs, and a bucket transfer process.
- session processing including recording and updating of session information indicating the state of the session to which the bucket belongs is performed, and the main processing is performed on packets for which the local class member is in charge of sub-processing.
- FIG. 1 is a block diagram showing the overall configuration of an embodiment according to the present invention.
- FIG. 2 is a block diagram showing a configuration example of the cluster member 11i.
- FIG. 3 is a diagram showing an example of the contents of the range holding unit 122.
- FIG. 4 is a diagram showing an example of the contents of the range holding unit 122.
- FIG. 5 is a diagram showing an example of the contents of the cluster member list 172.
- FIG. 6 is a flowchart showing a processing example when a bucket with a multicast address is received.
- FIG. 7 is a flowchart showing an example of the session processing performed by the session processing unit 13.
- FIG. 8 is a flowchart showing an example of the session processing performed by the session processing unit 13.
- FIG. 9 is a flowchart illustrating an example of the bucket transfer process performed by the packet transfer unit 15.
- FIG. 10 is a flowchart illustrating an example of an advertisement message transmission process performed by the advertisement message processing unit 18.
- FIG. 11 is a flowchart illustrating an example of the advertisement message receiving process performed by the advertisement message processing unit 18.
- FIG. 12 is a flowchart showing an example of the alive monitoring timer monitoring process performed by the failure recovery unit 20.
- FIG. 13 is a diagram illustrating an example of the contents of the range holding unit 122.
- FIG. 14 is a diagram illustrating an example of the contents of the range holding unit 122.
- FIG. 15 is a flowchart illustrating a processing example when the distributed processing control unit 17 receives a new member addition instruction.
- FIG. 16 is a flowchart illustrating a processing example when the distributed processing control unit 17 receives a timeout notification.
- FIG. 17 is a diagram for explaining an operation when a failed cluster member occurs.
- FIG. 18 is a diagram for explaining the operation when a new cluster member is added.
- FIG. 19 is a diagram for explaining another embodiment of the present invention.
- FIG. 20 is a block diagram for explaining a conventional technique.
- FIG. 21 is a block diagram for explaining another conventional technique. BEST MODE FOR CARRYING OUT THE INVENTION
- FIG. 1 is a block diagram showing a configuration example of an embodiment of a cluster system according to the present invention.
- the class system 1 of the present embodiment includes n (multiple) cluster members 1_1 to 11n. However, they are connected to adjacent IP nodes 2-1 to 2 _m via data links 3 _ 1 to 3 -m, respectively.
- FIG. 2 is a block diagram showing a configuration example of a cluster member l_i (l ⁇ i ⁇ n).
- Each cluster member 1-i has an interface section (IF section) 11-1 to 1 1-111 for each data link 3-1 to 3_m, a traffic distribution filter 12 and a session process.
- Unit 13 session information holding unit 14, packet transfer unit 15, routing table 16, distributed processing control unit 17, advertisement message processing unit 18, alive monitoring timer unit 19 And a failure recovery unit 20.
- the interface unit 1 1—j (1 ⁇ j ⁇ m) transmits the bucket advertisement message passed from the packet transfer unit 15 to the data link 3—j, and transmits the packet via the data link 3—j.
- the traffic distribution filter 1 2 includes a range holding unit 1 2 1.
- This range The holding unit 1 2 1 stores the range of the traffic that the local class member 1_i is currently in charge of the main processing (main processing range) and the range of the traffic that was in charge of the main processing in the past (old The main processing range), the range of traffic currently in charge of sub-processing (sub-processing range), and the range of traffic that was in charge of sub-processing in the past (former sub-processing range) are stored.
- the traffic distribution filter 12 determines whether to pass the packet based on each processing range (session processing unit). Pass it to 13) or decide to discard it. Note that the old main processing range and the old sub-processing range may not be stored in the range holding unit 121.
- Fig. 3 and Fig. 4 show examples of the contents of the range holding unit.
- the example in FIG. 3 shows that the buckets matching the main processing range m f i and the sub-processing ranges b f l, i to b f P, i are passed, and the other packets are discarded.
- the example in FIG. 4 matches the main processing range mf i ′, the old main processing range mfi, the sub-processing range bfl, i ′ to bf P, i ′ and the old sub-processing range bfl, i to bf P, i This indicates that the bucket should be passed and the other buckets should be discarded.
- the session processing unit 13 performs session processing on the packet that has passed through the traffic distribution filter 12. In this session processing, processing for identifying the session to which the packet belongs, processing for updating or deleting session information indicating the session state held in the session information holding unit 14, and processing for adding new information to the session information holding unit 14. This includes processing for registering invalid session information (when a session is opened) and processing for discarding an invalid bucket based on packet header information and session information.
- a session refers to a virtual communication path provided by an IP extension protocol and an upper layer protocol, and includes, for example, a TCP connection and a security association of IP sec.
- session state refers to these Refers to information unique to each connection.
- TCP connection it includes the connection status, sequence number, response number, and so on.
- IP sec SA it includes parameters of SA defined in RFC2401.
- the packet transfer unit 15 performs a routing table on a transmission route of a packet passed from the session processing unit 13 or a packet received by the interface unit 11-j and having a destination other than the data link multicast address. It has the function of determining based on the content of 16 and the destination of the above packet, and forwarding it to the next hop node.
- the alive monitoring timer section 19 is provided with an alive monitoring timer (not shown) for each of the class members 11 to 11n constituting the class configuration system 1. Each alive monitoring timer enters a timeout state when a predetermined time Tout elapses after resetting.
- the advertisement message processing unit 18 has a function of transmitting an advertisement message to other class members in the cluster system 1 every predetermined time Ts (Ts ⁇ Tout), and the other class members 1 ⁇ k ( l ⁇ k ⁇ n), the function to reset the alive monitoring timer for the cluster member 1-k provided in the alive monitoring timer section 19 when receiving an advertisement message from It has a function to output a new member addition instruction to the distributed processing controller 17 when it detects that a new class member has been added.
- the failure recovery unit 20 constantly monitors the status of the alive monitoring timer for each of the cluster members 1_1 to 1-n provided in the alive monitoring timer unit 19, and detects the cluster member in which a failure has occurred. Includes the function to detect and the identifier of the class member that failed to the distributed processing control unit 17 when detecting the alive monitoring event that timed out (when detecting the cluster member in which the failure occurred) Determines whether it is necessary to take over the time-out notification function or the main processing performed by the failed cluster member, and if it is necessary, retains the range in the traffic distribution filter 12 It has a function to rewrite the contents of the section 121.
- the distributed processing control unit 17 includes an identifier storage unit 171, a class member list 172, and a range determination rule storage unit 173.
- the identifier holding unit 171 stores an identifier (for example, IDi) unique to the own class member 11i.
- the class evening member list 172 includes a number indicating the position of each of the cluster members 111 to 111 constituting the class evening system 1 in the class evening member list 172, an identifier, A range in charge of processing (main processing range) and a range currently in charge of sub-processing (sub-processing range) are set.
- FIG. 5 shows an example of the cluster member list 172.
- the cluster member 11 i having the number “i” and the identifier “ID i” is composed of the main processing for the packet matching the main processing range m fi and the sub-processing range bfl, i to bf P , i indicates that it is in charge of the sub-processing of the priority “1” to “P” for the packet that matches i.
- P can be any value (integer) as long as it satisfies 1 ⁇ P ⁇ (n-1).
- the range determination rule holding unit 173 includes a main processing range determination rule for determining a main processing range mfl to mfn to be assigned to each of the cluster members 111 to 1-n constituting the class system 1,
- the sub-processing range determination rules for determining the sub-processing ranges bfl, l to bfp, n to be assigned to each class member 111-111 are stored.
- the main processing range determination rule that sets the main processing ranges mfl to mfn of the cluster members of numbers “1” to “ ⁇ ” to 1-1 to 1— ⁇ , respectively, to “0” to “n ⁇ 1”.
- the main processing range mfl to mfn (“0” to “n—1”) is the IP source address of the packet. Indicates the remainder obtained by performing a commutative operation (for example, adding or multiplying both) on the IP destination address and dividing the operation result by the number of clusters “n”.
- the rule for determining the main processing range is not limited to this, and any rule that can determine the main processing range that satisfies the above equations (1) and (2) can be used. good.
- the sub-processing ranges bf P, l to bf P, n of the priority P assigned to each cluster member 1-1 to 11-n determined by the sub-processing range determination rule do not overlap with each other, and Must cover all traffic T to be processed.
- the main processing range mfi assigned to the cluster member 1-i and the sub-processing extent bfl, assigned to the priority class "1" to "P" assigned to the class evening member 1_i. i to bf P, i must not overlap.
- the sub-processing range determination rule must be able to determine the sub-processing range bf P, l to bf ⁇ , ⁇ that satisfies the following equations (3) to (5).
- bf P, l U bf P, 2 U... U bf P, n T... (3)
- the distributed processing control unit 17 is configured to receive a new member addition instruction from the advertisement message processing unit 18 or to perform a failure recovery.
- the main processing of each cluster member 111-111 is performed according to the main processing range determination rule and sub-processing range determination rule stored in the range determination rule storage unit 173. It has a function of determining the range and the sub-processing range, a function of updating the contents of the cluster member list 17 2, a function of updating the contents of the range holding unit 12 1 in the traffic distribution filter 12, and the like.
- the class member 11i can be realized by a computer.
- the class member 11i When the class member 11i is realized by a computer, the following is performed. Prepare a disk, semiconductor memory, and other recording media on which programs for class members are recorded.
- the computer reads the above program and controls its own operation, so that the interface section 111-1-1 m and traffic distribution Filter 12, Session processing unit 13, Packet transfer unit 15, Routing table 16, Distributed processing control unit 17, Advertising message processing unit 18, Alive monitoring timer unit 19, Failure recovery unit 20 Realize. [Explanation of Operation of Embodiment]
- the class members 1-1-1 to 11-n Upon receiving the multicast packet, the class members 1-1-1 to 11-n perform the processing shown in the flowchart of FIG.
- the traffic distribution filter 12 in the cluster member 1_i receives the multicast packet via the interface 1 1-a, first, it checks whether or not the packet matches the main processing range. (Step S51). If the packet matches the main processing range, the packet is added with the match type “main processing range” and passed to the session processing unit 13 (step S56).
- step S52 to S55 If it does not match the main processing range, it matches only the old main processing range, matches both the old main processing range and the sub-processing range, or matches only the sub-processing range. It is checked whether or not it matches the old sub-processing range (steps S52 to S55).
- the packet matches only the old main processing range, the packet is added with the match type “old main processing range” and passed to the session processing unit 13 (“Y E S” in step S 52, S 56).
- step S5 If both the old main processing range and the sub-processing range are matched, the match type “old main processing range, sub-processing range” is added and passed to the session processing unit 13 (step S5). In 3 "YE S", S 56).
- a match type “sub-processing range” is added and passed to the session processing unit 13 (“YES” in step S54, S56).
- the match type “old sub-processing range” is added and passed to the session processing unit 13 (“YE S”, S 56 in step S55) o No match in any range In this case, the packet is discarded ("NO" in step S55, S59). If the old main processing range and the old sub-processing range are not stored in the range holding unit 121, the old main processing range and the old sub-processing range do not match.
- step S57 When the packet to which the match type is added is passed from the traffic distribution filter 12, the session processing unit 13 performs session processing (step S57).
- the session processing performed in step S57 will be described in detail with reference to FIGS.
- the session processing unit 13 converts the packet into a session. Then, it is determined whether or not the request is made (step S602).
- the packet requests a session to be opened, the session information of the session to which the bucket belongs is registered in the session information holding unit 14 and the packet is processed (steps S603 and S604).
- step S604 a process of discarding an invalid packet is performed (a bucket not discarded is passed to the packet transfer unit 15).
- the session information holding unit 14 stores the session to which the bucket belongs based on the header information of the packet. It is checked whether or not session information exists (step S605).
- the bucket is discarded (step S606), and if it exists (the self-cluster member belongs to the bucket to which the packet belongs).
- Session opening process If the packet is a packet that requests session termination, the process updates the session information and processes the bucket (steps S603 and S604). Then, a process of deleting the session information of the session to which the bucket belongs stored in the session information holding unit 14 is performed.
- step S607 If a packet to which the match type “old main processing range” is added is passed from the traffic distribution filter 12 (“YES” in step S607), whether the packet requests a session establishment Determine whether or not (step S607)
- step S If it is requested to open a session, discard it (step S
- step S610 it is checked whether or not the session information of the session to which the packet belongs is stored in the session information holding unit 14 (step S610).
- step S609 If the corresponding session information does not exist, the packet is discarded (step S609). If the corresponding session information exists, the corresponding session information stored in the session information holding unit 14 is updated, and Processing such as discarding the illegal bucket is performed (steps S611 and S612).
- the session processing unit 13 When a bucket to which the match type “sub-processing range” is added is passed from the traffic dispersion filter 12 (“YES” in step S701 of FIG. 8), the session processing unit 13 first It is determined whether or not the packet is for requesting the establishment of a session (step S702).
- the session information of the bucket is registered in the session information holding unit 14 and the bucket is discarded (steps S703 and S704).
- step S705 it is checked whether or not the session information of the session to which the packet belongs is stored in the session information holding unit 14 (step S705).
- the session information is updated, and the bucket is discarded (steps S703 and S704).
- the packet is discarded (step S706).
- step S707 If a packet to which the match type “old sub-processing range” is added is passed from the traffic distribution filter 12 (“YES” in step S707), it is determined whether or not the bucket requests session establishment. Is checked (step S708). If it is a packet requesting the establishment of a session, discard processing is performed (step S709). Otherwise, it is determined whether or not the session information corresponding to the session information is stored in the session information holding unit 14. Check (step S710).
- step S709 If the corresponding session information does not exist, the packet is discarded (step S709), and if there is, the session information is updated and the bucket is discarded (steps S711, S712).
- step S707 If a packet to which the match type “old main processing range, sub-processing range” is added is passed from the traffic distribution filter 12 (“NO” in step S707), the packet requests to open a session. A check is made to determine whether or not (step S713).
- the packet is a request to open a session
- the session information of the session to be opened is registered in the session information holding unit 14, and the bucket is discarded (steps S714 and S715).
- step S716 if the packet does not request a session establishment, it is checked whether or not the corresponding session information is registered in the session information holding unit 14 (step S716).
- step S715 If it has not been registered, the bucket is discarded (step S715). If it has been registered, the session information registered in the session information holding unit 14 is updated, and the packet is updated. The packet is transferred to the packet transfer unit 15 (steps S717 and S718).
- the packet transfer unit 15 executes the packet transfer processing ( Step S 58).
- the routing table 16 is determined based on the IP address of the bucket.
- the transmission route is obtained by searching, and the packet is transmitted via the interface section corresponding to the obtained transmission route (steps S81 and S82).
- the packet transfer unit 15 performs the same processing on a packet received via the interface units 11-1 to 11-m and not addressed to the data link multicast address.
- the advertisement message processing unit 18 in the cluster member 11 i sends the cluster member list every time the advertisement message transmission timing comes (“YES” in step S 91). All identifiers stored in 1 7 2 (identifiers of class members that recognize that their own class member 1 i exists in the cluster system 1), and identifier holder 1 7 1 (Step S92), and sends an advertisement message including them to the other class members in the class system 1 (step S92). S93).
- the advertisement message processing unit 18 transmits the advertisement message using a predetermined multicast or broadcast address other than the cluster IP address.
- step S91 it is determined whether or not the transmission timing has been reached based on whether or not a predetermined time Ts has elapsed since the previous transmission of the advertisement message. This time is shorter than the timer time-out time Tout. That is, the relationship of Ts ⁇ Tout is satisfied. Also, the cluster member list 172 in the class member newly added to the cluster system 1 contains only the identifier of the own cluster member. [Advertisement message receiving process]
- an advertisement message processing unit 18 in cluster member 1—i receives an advertisement message from another cluster member in class system 1 (for example, cluster member l_n)
- the advertisement message processing unit 18 in FIG. As shown in the chart, the source cluster member included in the advertising message It is determined whether the identifiers 1-n are registered in the cluster member list 172 (step S101).
- step S 101 If it is determined that the identifier of the class member 11 is registered in the cluster member list 172 (“YES” in step S 101), the cluster monitoring function provided in the alive monitoring timer unit 19 is performed. Reset the alive monitoring timer (not shown) for member 111, and then pass all identifiers included in the advertisement message to distributed processing controller 17 (steps S102, S103) 0
- step S101 if it is determined that a new cluster member has not been registered ("NO" in step S101), it is recognized that a new cluster member has been added to the cluster system 1, and the distribution processing control unit 17 adds a new member.
- the instruction is output (step S104).
- This new member addition instruction includes the identifier of the cluster member 110 n of the transmission source of the advertisement message.
- the advertisement message processing unit 18 generates an alive monitoring event for the new class evening members 1-n in the alive monitoring event unit 19 (step S105).
- Each of the alive monitoring timers provided in the alive monitoring timer unit 19 is set to a time-out state when a predetermined time Tout elapses after being reset.
- the failure recovery unit 20 in the class evening member 11 i is alive and alive monitoring for each class evening member provided in the alive monitoring timer unit 19 as shown in FIG. Attention is paid to one of the timers, and it is checked whether or not the alive monitoring timer has timed out (steps S111, S112).
- step S112 If the time-out state has not been reached ("NO" in step S112), attention is paid to the next alive monitoring event (step S111).
- step S112 the alive monitoring timer (for example, the alive monitoring timer for the cluster member 111) that is currently focused on is in a timeout state ("YES" in step S112), the own class Evening member 1—i is in charge of the sub-processing with priority “1” for the traffic that cluster member 1-n is in charge of the main processing. It is determined whether or not there is (step S113). This determination is made with reference to the cluster member list 172.
- step S113 If it is in charge of the sub-processing of the priority “1” (“YE S” in step S113), the contents of the range holding unit 121 in the traffic distribution filter 12 are updated, and the class members are updated. Even for the range of traffic for which n was in charge of the main processing, the own cluster member 1_i is also in charge of the main processing (step S114).
- the sub-process in charge of cluster member 1-n includes its own class. Evening member 1—determines whether there is a sub-process that has the same traffic range as the sub-process that is assigned to i and has a higher priority than the sub-process that is assigned to the local cluster member 11-i (step S 117).
- step S117 If such a sub-process exists ("YE S" in step S117), the priority of the corresponding sub-process range is increased by one (step S118).
- b f2, l indicates the same range as the traffic range in which class members 1-n were in charge of the sub-processing with priority "1".
- step S115 the failure recovery unit 20 notifies the distributed processing control unit 17 of a timeout.
- This timeout notification includes the identifier of the cluster member 1-n corresponding to the alive monitoring event that has timed out. So After that, the failure recovery unit 20 deletes the alive monitoring timer for the cluster member 111 from the alive monitoring timer unit 19 (step S116), and performs the processing of step S111 again.
- the distributed processing control unit 17 in the cluster member 11 i performs the processing shown in the flowchart of FIG. It is assumed that the new member addition instruction includes the identifier of the cluster member 1 1 X as the identifier of the newly added class member.
- the distributed processing control unit 17 uses the main processing range determination rule and the sub-processing range determination rule stored in the range determination rule holding unit 17 3 to determine the main process of each class member in the cluster system 1.
- the processing range and the sub-processing range are determined (Step S1311). In other words, the main processing range and sub-processing range of each cluster member whose identifier is stored in the class member list 17 2 and the added cluster member 1 — X are determined.
- the distributed processing control unit 17 sorts the information (identifier, main processing range, and sub-processing range) of each cluster member by using the identifier as a key, assigns a number in the sorting order, and classifies the cluster member list. 2 (step S1 32).
- the distributed processing control unit 17 determines whether or not the class system configuration recognized by its own class member 1-i is the same as the cluster system configuration recognized by the other cluster members. It is determined (step S133).
- step S103 of FIG. This determination is made based on all the identifiers in the advertisement message input from the advertisement message processing unit 18 in step S103 of FIG. That is, an advertisement message is sent from each cluster member that recognizes that the own cluster member 11i exists in the cluster system 1, and an advertisement message sent from each of the above cluster members. If all the identifiers in the class match the identifiers stored in the class member list 1 72 in the own class member 1—i, the class system configuration recognized by all class members Judge that they are the same. In step S133, if the cluster system recognizes that all cluster members recognize the same class configuration, the distributed processing control unit 17 executes the range holding unit 1 2 1 in the traffic distribution filter 12 Is updated (step S1 3 4).
- the distributed processing control unit 17 performs the processing shown in the flowchart of FIG. It is assumed that the timeout notification includes a child.
- the distributed processing control unit 17 deletes the entry of the cluster member 111 from the distributed processing control unit 17 (step S144).
- each cluster member whose identifier is stored in the cluster member list 172 is determined.
- the main processing range and the sub-processing range are determined and determined.
- the cluster member list 172 is updated based on the main processing range and the sub-processing range (steps S142, 144).
- step S144 determines whether or not the cluster system configuration recognized by all cluster members is the same (step S144), and if so, the traffic distribution is performed.
- the contents of the range holding section 122 in the filter 12 are updated (step S145).
- step S145 the same processing as in step S134 described above is performed.
- the cluster system 1 has four class evening members 1-1 to 1-14 Assume that T is configured and all traffic to be processed by the cluster system 1 is T. Also, the main processing ranges mfl to mf4 and the sub-processing ranges bfl, l to bf1, 4 shown in FIGS. 17A and 17B are set for each of the cluster members 1_1 to 1-4. And That is, the sub-process with priority “1” related to the main process assigned to the cluster members 1-1, 1—2, 1 ⁇ 3, and 1-4 is assigned to the class members 1-4, 1 ⁇ 1. , 1-2, 1-3 are in charge. At this time, it is assumed that the old main processing range and the old sub-processing range have not been set.
- the class members 1-1, 1-1-2 discard the packet because the above packet does not match any of the main processing range, the old processing range, the sub-processing range, and the old sub-processing range (see step 55 in FIG. 6). All "NO”, S59).
- the cluster members 13 and 13 execute session processing for the above-mentioned packet ("YE S" in step S54, S56, S57). .
- the cluster members 114 execute session processing for the above-mentioned packet (“YES” in step S 51, S 56, S 57).
- the cluster members 1-3 execute the time-out notification receiving process shown in the flowchart of FIG. Since the other cluster members 111 and 112 become "NO" in steps S113 and S117 in FIG. 12, respectively, the time chart notification reception processing shown in the flowchart in FIG. 16 is executed. I do.
- each of the cluster members 111 to 113 operating normally has the main processing range mf l ′ shown in FIGS. 17C and 17D.
- the main processing range mfl-mf3 and the sub-processing range bfl, l-bf1,3 are the old main processing range and the old sub-processing range of the cluster members 1-1-1-13.
- the traffic distribution filter 12 in the cluster member 1_1 discards the bucket because the packet does not match any of the main processing range, the old main processing range, the sub-processing range, and the old sub-processing range (FIG. 6). NO in step S55, S59).
- the traffic distribution filter 12 in the class member 1 _2 adds the match type “sub-processing range” to the session because the above-mentioned packet matches only the sub-processing range bf 1, 2 ′. It is passed to the processing unit 13 ("YE S" in step S54 of FIG. 6, S56).
- the session processing unit 13 stores the session information of the packet in the session information holding unit 14 because the packet is not a packet for requesting a session establishment. It is checked whether it is (“YE S” in step S701 of FIG. 8, “NO” in S702, S705). In this example, since the corresponding session information is not stored, the packet is discarded ("NO" in S705, S706).
- the session processing unit 13 stores the session information related to the above-mentioned bucket in the session information holding unit 14 because the packet is not a packet requesting a session establishment. (“YES” in step S601, “N ⁇ ” in S602, S605) in FIG. 7 o
- the class members 113 are cluster members 1-4. Before the occurrence of the trouble, the sub-process was in charge of the packet belonging to the session ⁇ , so that the session information holding unit 14 has the corresponding session information.
- the session processing unit 13 updates the corresponding session information stored in the session information holding unit 14 (step S603), and further passes the bucket to the session processing unit 13 (step S604). ) 0
- the traffic distribution filter 12 in the class member 11 sends the session processing unit 13 with the match type "main processing range” added. "YES” in step S51 of step 6, S56).
- the session processing unit 13 registers the session information relating to the session / 3 in the session information holding unit 14 because the packet is a packet requesting to open a session.
- the packet is transferred to the packet transfer unit 15 ("YES" in S601, "YES” in S602, S603, S604 in FIG. 7).
- the traffic distribution filter 12 in the class members 1 and 2 passes the packet to the session processing unit 13 with the match type “old main processing range” added. "YES” in S52 of S6, S56).
- the session processing unit 13 discards the packet because it requests a session to be opened (“YE S” in step S607 in FIG. 7).
- “YE S” in S608, S609) the cluster member 111 is in charge of the main processing for the packet, the cluster member 1_2 discards the bucket.
- the traffic distribution filter 12 in the cluster member 13 adds the match type “sub-processing range” to the session processing unit 13 because the packet matches only the sub-processing range bf 1,3 ′ (see FIG. 6). "YES” in step S54, S56).
- the session processing unit 13 registers the session information on the session 3 in the session information holding unit 14 because it requests the session to be opened. At the same time, the packet is discarded ("YE S" in step S701 in FIG. 8, "
- the above is the operation when a failure occurs in the cluster members 1 to 4 in the class system 1.
- the operation when a new cluster member is added is as follows.
- the cluster system 1 is composed of three cluster members 1_1 to 1_3, and that all traffic to be processed by the class system 1 is T.
- the main processing ranges mfl to mf3 and the sub-processing ranges bfl, l to bf1, 3 shown in FIGS. 18A and 18B are set for each cluster member 1-1 to 1-3.
- the cluster members 1-3, 1-1, and 1-2 are in charge of the sub-processing of priority "1" related to the main processing assigned to the class members 1-1, 1-2, 1-3. are doing.
- the old main processing range and old sub-processing range have not been set.
- the cluster member 111 discards the packet because the packet does not match any of the main processing range, the old main processing range, the sub-processing range, and the old sub-processing range ("N" in step S55 in FIG. 6). ⁇ ", S 59).
- Cluster members 1 and 2 indicate that the above packet matches only sub-processing range bf 1,2 Therefore, execute the session processing (“YE S” in step S54, S56, class evening members 1-3) execute the session processing because the above packet matches the main processing range mf3 (step S54). "YE S” in S51, S56, S57) In such a state, if the cluster members 1-4 are added, the range holding unit 12 in each of the cluster members 1-1 to 1-4 is added. The contents of 1 are updated, and the main processing range, old main processing range, sub-processing range, and old sub-processing range assigned to each class evening member 111 to 114 are updated (step S13 in Fig. 15). 1 to S 134).
- each of the cluster members 1-1 to 1-4 has a main processing range mf 1 'to mf4' and a sub processing range bf 1,1 'to bf 1,4 shown in FIGS. 'Is set.
- the main processing range mfl to mf3 and the sub-processing range bfl, l to bf1,3 are the old main processing range and the old sub-processing range of the cluster members 1-1 to 113.
- each of the cluster members 1-1 to 1-4 receives the packet.
- the traffic distribution filter 12 in the class member 1-11 discards the packet because the packet does not match any of the main processing range, the old main processing range, the sub-processing range, and the old sub-processing range ( "NO” in step S55 of FIG. 6, S59).
- the traffic distribution filter 12 in the cluster members 1 and 2 adds the match type "old sub-processing range" to the session processing unit 13 because the bucket matches only the old sub-processing range bf1,2. Hand over (“YE S" in step S55 of FIG. 6, S56).
- the session processing unit 13 When a packet to which the match type “old sub-processing range” is added is passed, the session processing unit 13 does not request the packet to open a session, and the corresponding session information exists in the session information holding unit 14. Therefore, after updating the session information, the packet is discarded (“YES” in step S707, “N ⁇ ” in S708, “N ⁇ ” in S710, S709 in FIG. 8).
- the traffic distribution filter 1 2 in the cluster member 1-3 matches the bucket with both the old main processing range mi3 and the sub-processing range bf 1,3 '. "Old main processing range, sub-processing range” is added and passed to the session processing unit 13 ("YE S", S56 in step S53 in FIG. 6) o
- the session processing unit 13 When the packet added with the match type “old main processing range and sub-processing range” is passed from the traffic distribution filter 12, the session processing unit 13 does not request the session to be opened, and Since there is session information corresponding to the session information holding unit 14, after updating the session information, the packet is transferred to the bucket transfer unit 15 (“N ⁇ ” in step S713 and S716 in FIG. 8). "YES”, S717, S718).
- the above packet matches the main processing range mf 4 ′, so the match type “main processing range” is added and passed to the session processing unit 13 (FIG. 6).
- step S51 "YE S", S56).
- the session processing unit 13 does not request that the packet open a session, and the session information corresponding to the session information holding unit 14 The bucket is discarded because it does not exist ("YE S" in step S601, "N ⁇ ” in S602, “N ⁇ ” in S605, S606 in FIG. 7).
- the above packet matches both the old main processing range mfl and the sub processing range bf 1,1 ', so the match type is “old main processing range, sub processing range”. (“YE S” in step S53 of FIG. 6, S56) o
- the session processing unit 13 updates the session information (session ⁇ 5 After registering the session information in the session information holding unit 14), the packet is discarded (“YE S”, S714, S715 in step S713 in FIG. 8).
- the traffic distribution filter 12 in the cluster members 1 and 2 Since it matches the main processing range mf 2 ′, a match type “main processing range” is added and passed to the session processing unit 13 (“YES” in step S 51 of FIG. 6, S 56).
- the session processing unit 13 sends the session information to the session information holding unit 14 because it is a packet requesting the establishment of the session (5).
- the packet is transferred to the packet transfer unit 15 (“YE S”, S 603, and S 604 in steps S 601 and S 602 in FIG. 7).
- the traffic distribution filter 12 in the cluster member 13 adds the match type “old sub-processing range” and passes it to the session processing unit 13 (step in FIG. 6). "YES” in S55, S56).
- the session processing unit 13 discards the packet because it requests the establishment of session ⁇ 5 ("YE S" in steps S707 and S708 in FIG. 8, S709).
- the traffic distribution filter 12 in the cluster member 1-4 discards the bucket because the bucket does not match any of the main processing range, the old main processing range, the sub-processing range, and the old sub-processing range (see FIG. 6). "N ⁇ " in step S55, S59).
- the relevant session is selected from the old main processing range and the old sub-processing range stored in the range holding unit 121 in the traffic distribution file 12.
- the old main processing range and old sub-processing range in which (the session in which the opening process was performed in the own cluster member) has completely disappeared may be deleted.
- the old main processing range and the old sub-processing range are valid only when there is a related session within the range, and are meaningless when there is no related session. . By doing so, the number of processing ranges to be compared in the traffic distribution filter 12 is reduced, so that the processing speed can be improved.
- the class member performing the main process when the class member performing the main process fails, the class member performing the sub process performs the corresponding sub process in addition to its own main process.
- the processing is switched to the main processing (failover processing). If traffic processing is subdivided and sessions in progress disappear, the load on each member is again averaged, but in the short term after failover, processing is skewed to specific cluster members.
- the division of the range of traffic handled by each class member is finely defined as follows, thereby reducing the uneven processing load after failover.
- each cluster member is responsible for (n-1) of these, as the main processing. Also, each cluster member is responsible for (n-1) divisions in which any other cluster member is in charge of the main processing as a sub-processing. However, this division should be performed from the range in which different cluster members are all performing main processing.
- the sub-processing range b f i of the cluster member 1 _ i is determined so that
- Figure 19 shows an example of assignment.
- the first class members, the first failure recovery method and the first program according to the present invention session information is exchanged between cluster members. Without this, it is possible to perform failure recovery processing. As a result, session information was exchanged between cluster members. As compared with the conventional method, it is possible to perform the failure recovery process more quickly and to enhance the scalability. The reason is that the cluster member that is in charge of the sub-process performs the recording and updating process of the session information, and if a failure occurs in the cluster member that is in charge of the main process, it takes charge of the sub-process. This is because one class member takes over the processing using the session information recorded in the own cluster member.
- each class member can be replaced when a failed cluster member occurs or a new class member is added. It is possible to change the main processing range and sub-processing range that the member is responsible for. The reason is that each cluster member determines a new main processing range and sub-processing range based on the range determination rule, and sets the previous main processing range and sub-processing range as the old main processing range and old sub-processing range. This is because they do so.
Landscapes
- Engineering & Computer Science (AREA)
- Computer Security & Cryptography (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
- Computer And Data Communications (AREA)
Abstract
Description
Claims
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2005516288A JP4458289B2 (ja) | 2003-12-17 | 2004-11-29 | クラスタシステム、クラスタメンバ、故障復旧方法及びプログラム |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2003419005 | 2003-12-17 | ||
JP2003-419005 | 2003-12-17 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2005060187A1 true WO2005060187A1 (ja) | 2005-06-30 |
Family
ID=34675195
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2004/018090 WO2005060187A1 (ja) | 2003-12-17 | 2004-11-29 | クラスタシステム、クラスタメンバ、故障復旧方法及びプログラム |
Country Status (4)
Country | Link |
---|---|
US (1) | US7443787B2 (ja) |
JP (1) | JP4458289B2 (ja) |
CN (1) | CN1914862A (ja) |
WO (1) | WO2005060187A1 (ja) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2007220103A (ja) * | 2006-02-16 | 2007-08-30 | Internatl Business Mach Corp <Ibm> | 結合セッション環境におけるセッションのクラスタのためのセッション情報の統合方法、システム、およびプログラム |
CN100411460C (zh) * | 2006-07-20 | 2008-08-13 | 华为技术有限公司 | 中继移动交换中心离开组呼的方法 |
CN101192061B (zh) * | 2006-11-29 | 2010-07-07 | 国际商业机器公司 | 用于同步控制器固件下载的装置及方法 |
JP2011100323A (ja) * | 2009-11-06 | 2011-05-19 | Hitachi High-Technologies Corp | 画像データ配信方法及び検査装置 |
WO2011148493A1 (ja) * | 2010-05-27 | 2011-12-01 | 富士通株式会社 | パケット通信装置及びパケット転送方法 |
WO2012060276A1 (ja) * | 2010-11-01 | 2012-05-10 | かもめエンジニアリング株式会社 | アクセス制御方法、アクセス制御装置およびアクセス制御プログラム |
JP2012222694A (ja) * | 2011-04-12 | 2012-11-12 | Nippon Telegr & Teleph Corp <Ntt> | マルチキャスト配信システム、配信ルータ、および、マルチキャスト配信方法 |
Families Citing this family (36)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7969991B2 (en) * | 2007-04-23 | 2011-06-28 | Mcafee, Inc. | Session announcement system and method |
US8547844B2 (en) * | 2007-07-10 | 2013-10-01 | Telefonaktiebolaget L M Ericsson (Publ) | System and method for balancing IP gateway services |
WO2011120214A1 (zh) * | 2010-03-29 | 2011-10-06 | 华为技术有限公司 | 集群路由器和集群路由方法 |
US9569439B2 (en) | 2011-10-31 | 2017-02-14 | Elwha Llc | Context-sensitive query enrichment |
US10552581B2 (en) | 2011-12-30 | 2020-02-04 | Elwha Llc | Evidence-based healthcare information management protocols |
US10679309B2 (en) | 2011-12-30 | 2020-06-09 | Elwha Llc | Evidence-based healthcare information management protocols |
US10559380B2 (en) | 2011-12-30 | 2020-02-11 | Elwha Llc | Evidence-based healthcare information management protocols |
US20130173295A1 (en) | 2011-12-30 | 2013-07-04 | Elwha LLC, a limited liability company of the State of Delaware | Evidence-based healthcare information management protocols |
US10475142B2 (en) | 2011-12-30 | 2019-11-12 | Elwha Llc | Evidence-based healthcare information management protocols |
US10528913B2 (en) | 2011-12-30 | 2020-01-07 | Elwha Llc | Evidence-based healthcare information management protocols |
US10340034B2 (en) | 2011-12-30 | 2019-07-02 | Elwha Llc | Evidence-based healthcare information management protocols |
US9781664B2 (en) | 2012-12-31 | 2017-10-03 | Elwha Llc | Cost-effective mobile connectivity protocols |
US9876762B2 (en) | 2012-12-31 | 2018-01-23 | Elwha Llc | Cost-effective mobile connectivity protocols |
US9451394B2 (en) | 2012-12-31 | 2016-09-20 | Elwha Llc | Cost-effective mobile connectivity protocols |
US9980114B2 (en) | 2013-03-15 | 2018-05-22 | Elwha Llc | Systems and methods for communication management |
US9832628B2 (en) | 2012-12-31 | 2017-11-28 | Elwha, Llc | Cost-effective mobile connectivity protocols |
US9635605B2 (en) | 2013-03-15 | 2017-04-25 | Elwha Llc | Protocols for facilitating broader access in wireless communications |
US9713013B2 (en) | 2013-03-15 | 2017-07-18 | Elwha Llc | Protocols for providing wireless communications connectivity maps |
US8965288B2 (en) | 2012-12-31 | 2015-02-24 | Elwha Llc | Cost-effective mobile connectivity protocols |
US9866706B2 (en) | 2013-03-15 | 2018-01-09 | Elwha Llc | Protocols for facilitating broader access in wireless communications |
US9807582B2 (en) | 2013-03-15 | 2017-10-31 | Elwha Llc | Protocols for facilitating broader access in wireless communications |
US9706382B2 (en) | 2013-03-15 | 2017-07-11 | Elwha Llc | Protocols for allocating communication services cost in wireless communications |
US9693214B2 (en) | 2013-03-15 | 2017-06-27 | Elwha Llc | Protocols for facilitating broader access in wireless communications |
US9813887B2 (en) | 2013-03-15 | 2017-11-07 | Elwha Llc | Protocols for facilitating broader access in wireless communications responsive to charge authorization statuses |
US9781554B2 (en) | 2013-03-15 | 2017-10-03 | Elwha Llc | Protocols for facilitating third party authorization for a rooted communication device in wireless communications |
US9843917B2 (en) | 2013-03-15 | 2017-12-12 | Elwha, Llc | Protocols for facilitating charge-authorized connectivity in wireless communications |
US9706060B2 (en) | 2013-03-15 | 2017-07-11 | Elwha Llc | Protocols for facilitating broader access in wireless communications |
US9596584B2 (en) | 2013-03-15 | 2017-03-14 | Elwha Llc | Protocols for facilitating broader access in wireless communications by conditionally authorizing a charge to an account of a third party |
US9805208B2 (en) | 2013-09-30 | 2017-10-31 | Elwha Llc | Mobile device sharing facilitation methods and systems with recipient-dependent inclusion of a data selection |
US9813891B2 (en) | 2013-09-30 | 2017-11-07 | Elwha Llc | Mobile device sharing facilitation methods and systems featuring a subset-specific source identification |
US9838536B2 (en) | 2013-09-30 | 2017-12-05 | Elwha, Llc | Mobile device sharing facilitation methods and systems |
US9740875B2 (en) | 2013-09-30 | 2017-08-22 | Elwha Llc | Mobile device sharing facilitation methods and systems featuring exclusive data presentation |
US9774728B2 (en) | 2013-09-30 | 2017-09-26 | Elwha Llc | Mobile device sharing facilitation methods and systems in a context of plural communication records |
US9826439B2 (en) | 2013-09-30 | 2017-11-21 | Elwha Llc | Mobile device sharing facilitation methods and systems operable in network equipment |
US10572293B2 (en) * | 2017-12-15 | 2020-02-25 | Nicira, Inc. | Node in cluster membership management protocol |
US10476744B2 (en) | 2017-12-15 | 2019-11-12 | Nicira, Inc. | Coordinator in cluster membership management protocol |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP3062155B2 (ja) * | 1998-07-31 | 2000-07-10 | 三菱電機株式会社 | 計算機システム |
JP2002351760A (ja) * | 2001-05-30 | 2002-12-06 | Mitsubishi Electric Corp | サーバ負荷分散装置、サーバ負荷分散方法およびその方法をコンピュータに実行させるプログラム |
JP3366804B2 (ja) * | 1995-09-08 | 2003-01-14 | 株式会社日立製作所 | インタネットワーク装置 |
JP2003131961A (ja) * | 2001-07-09 | 2003-05-09 | Hitachi Ltd | ネットワークシステム及び負荷分散方法 |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9197599B1 (en) * | 1997-09-26 | 2015-11-24 | Verizon Patent And Licensing Inc. | Integrated business system for web based telecommunications management |
US7480606B2 (en) * | 1998-08-31 | 2009-01-20 | Versity Design, Inc. | VCD-on-demand system and method |
US6006259A (en) | 1998-11-20 | 1999-12-21 | Network Alchemy, Inc. | Method and apparatus for an internet protocol (IP) network clustering system |
US6078957A (en) | 1998-11-20 | 2000-06-20 | Network Alchemy, Inc. | Method and apparatus for a TCP/IP load balancing and failover process in an internet protocol (IP) network clustering system |
WO2002001410A1 (en) * | 2000-06-26 | 2002-01-03 | International Business Machines Corporation | Data management application programming interface for a parallel file system |
US7080225B1 (en) * | 2002-12-10 | 2006-07-18 | Emc Corporation | Method and apparatus for managing migration of data in a computer system |
-
2004
- 2004-11-29 JP JP2005516288A patent/JP4458289B2/ja not_active Expired - Fee Related
- 2004-11-29 CN CNA2004800414711A patent/CN1914862A/zh active Pending
- 2004-11-29 WO PCT/JP2004/018090 patent/WO2005060187A1/ja active Application Filing
- 2004-12-17 US US11/016,484 patent/US7443787B2/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP3366804B2 (ja) * | 1995-09-08 | 2003-01-14 | 株式会社日立製作所 | インタネットワーク装置 |
JP3062155B2 (ja) * | 1998-07-31 | 2000-07-10 | 三菱電機株式会社 | 計算機システム |
JP2002351760A (ja) * | 2001-05-30 | 2002-12-06 | Mitsubishi Electric Corp | サーバ負荷分散装置、サーバ負荷分散方法およびその方法をコンピュータに実行させるプログラム |
JP2003131961A (ja) * | 2001-07-09 | 2003-05-09 | Hitachi Ltd | ネットワークシステム及び負荷分散方法 |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2007220103A (ja) * | 2006-02-16 | 2007-08-30 | Internatl Business Mach Corp <Ibm> | 結合セッション環境におけるセッションのクラスタのためのセッション情報の統合方法、システム、およびプログラム |
US8626722B2 (en) | 2006-02-16 | 2014-01-07 | International Business Machines Corporation | Consolidating session information for a cluster of sessions in a coupled session environment |
CN100411460C (zh) * | 2006-07-20 | 2008-08-13 | 华为技术有限公司 | 中继移动交换中心离开组呼的方法 |
CN101192061B (zh) * | 2006-11-29 | 2010-07-07 | 国际商业机器公司 | 用于同步控制器固件下载的装置及方法 |
JP2011100323A (ja) * | 2009-11-06 | 2011-05-19 | Hitachi High-Technologies Corp | 画像データ配信方法及び検査装置 |
JP5423886B2 (ja) * | 2010-05-27 | 2014-02-19 | 富士通株式会社 | パケット通信装置及びパケット転送方法 |
WO2011148493A1 (ja) * | 2010-05-27 | 2011-12-01 | 富士通株式会社 | パケット通信装置及びパケット転送方法 |
US9088500B2 (en) | 2010-05-27 | 2015-07-21 | Fujitsu Limited | Packet communication apparatus and packet transfer method |
JPWO2012060276A1 (ja) * | 2010-11-01 | 2014-05-12 | かもめエンジニアリング株式会社 | アクセス制御方法、アクセス制御装置およびアクセス制御プログラム |
JP5538560B2 (ja) * | 2010-11-01 | 2014-07-02 | かもめエンジニアリング株式会社 | アクセス制御方法、アクセス制御装置およびアクセス制御プログラム |
US8850047B2 (en) | 2010-11-01 | 2014-09-30 | Kamome Engineering, Inc. | Access control method, access control apparatus, and access control program |
WO2012060276A1 (ja) * | 2010-11-01 | 2012-05-10 | かもめエンジニアリング株式会社 | アクセス制御方法、アクセス制御装置およびアクセス制御プログラム |
JP2012222694A (ja) * | 2011-04-12 | 2012-11-12 | Nippon Telegr & Teleph Corp <Ntt> | マルチキャスト配信システム、配信ルータ、および、マルチキャスト配信方法 |
Also Published As
Publication number | Publication date |
---|---|
US20050135343A1 (en) | 2005-06-23 |
US7443787B2 (en) | 2008-10-28 |
CN1914862A (zh) | 2007-02-14 |
JPWO2005060187A1 (ja) | 2007-07-12 |
JP4458289B2 (ja) | 2010-04-28 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2005060187A1 (ja) | クラスタシステム、クラスタメンバ、故障復旧方法及びプログラム | |
US10284389B2 (en) | High availability for distributed network services in an extended bridge | |
US9948553B2 (en) | System and method for virtual network-based distributed multi-domain routing control | |
CN101471885B (zh) | 具有状态同步的集群的虚拟组播路由 | |
CN104488238B (zh) | 用于网络环境中集群链路聚合控制的系统和方法 | |
US7859992B2 (en) | Router redundancy in data communication networks | |
US8009684B2 (en) | High capacity ring communication network | |
US7751394B2 (en) | Multicast packet relay device adapted for virtual router | |
EP2731313B1 (en) | Distributed cluster processing system and message processing method thereof | |
JP2016535904A (ja) | ミドルウェアおよびアプリケーションの実行のためにエンジニアド・システムにおいてデータサービスを提供するためのシステムおよび方法 | |
JP3888209B2 (ja) | マルチキャスト通信装置およびシステム | |
CN102263646B (zh) | 交换机的分布式控制平面内的多播 | |
CN104106244A (zh) | 控制装置、通信系统、通信方法和程序 | |
US20060050643A1 (en) | Router for multicast redundant routing and system for multicast redundancy | |
US20140211797A1 (en) | Managing multicast distribution using multicast trees | |
CN1937569A (zh) | 具有冗余路径的局域网中的消息处理 | |
WO2021082812A1 (zh) | 报文的发送方法和第一网络设备 | |
JP4524686B2 (ja) | クラスタシステム及びクラスタメンバ並びにプログラム | |
CN104734877B (zh) | 一种获取配置服务器信息的方法、装置及系统 | |
WO2013141191A1 (ja) | 制御装置、通信システム、ノード制御方法及びプログラム | |
KR100476237B1 (ko) | 복수 방화벽에서의 효율적 부하 분산을 위한 비대칭 트래픽 처리 방법 | |
JP2005057693A (ja) | ネットワーク仮想化システム | |
JP3930445B2 (ja) | マルチキャスト通信方法、マルチキャスト通信用ルータ装置およびマルチキャスト通信用プログラム | |
JP2005286681A (ja) | 中継機器 | |
WO2024067367A1 (zh) | 用于确定节点的方法及装置 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 200480041471.1 Country of ref document: CN |
|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BW BY BZ CA CH CN CO CR CU CZ DE DK DM DZ EC EE EG ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NA NI NO NZ OM PG PH PL PT RO RU SC SD SE SG SK SL SY TJ TM TN TR TT TZ UA UG US UZ VC VN YU ZA ZM ZW |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): BW GH GM KE LS MW MZ NA SD SL SZ TZ UG ZM ZW AM AZ BY KG KZ MD RU TJ TM AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IS IT LU MC NL PL PT RO SE SI SK TR BF BJ CF CG CI CM GA GN GQ GW ML MR NE SN TD TG |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWE | Wipo information: entry into national phase |
Ref document number: 2005516288 Country of ref document: JP |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWW | Wipo information: withdrawn in national office |
Ref document number: DE |
|
122 | Ep: pct application non-entry in european phase |