CN101488918B - Multi-network card server access method and system - Google Patents
Multi-network card server access method and system Download PDFInfo
- Publication number
- CN101488918B CN101488918B CN2009100766005A CN200910076600A CN101488918B CN 101488918 B CN101488918 B CN 101488918B CN 2009100766005 A CN2009100766005 A CN 2009100766005A CN 200910076600 A CN200910076600 A CN 200910076600A CN 101488918 B CN101488918 B CN 101488918B
- Authority
- CN
- China
- Prior art keywords
- network interface
- interface card
- host exchange
- switch
- backup group
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
- Small-Scale Networks (AREA)
Abstract
The invention discloses an access method and system of multi-NIC server. The method comprises the steps of: composing a VRRP backup group by a plurality of three-tier switches, configuring the routing priority of the main switch higher than the routing priority of the accessory switch in the VRRP backup group; connecting the main network card in the multi-NIC server with the main switch in the VRRP backup group; connecting the spare network card with the accessory switch in the VRRP backup group; receiving the downlink data by the main switch having higher routing priority in the VRRP backup group, using the studied address information of the main network card to send the received downlink data to the main network card in the multi-NIC server; sending the uplink data to the main switch in the VRRP backup group according to the obtained virtual address information of the VRRP backup group by the network card ensured by load balancing algorithm in the multi-NIC server, and transmitting by the main switch. The method and system in the invention can improve the reliability of the multi-NIC server attaching to the network.
Description
Technical field
The present invention relates to the network communications technology, particularly relate to a kind of cut-in method and system of multi-network card server.
Background technology
Service application based on multi-network card server has appearred in progressively development along with network traffic at present, such as the multi-network card server that in data center (IDC) business, just has been widely used (being generally two network card servers).Disposed a plurality of network interface cards (NIC) in the multi-network card server, these a plurality of network interface cards are bundled into a group, a shared IP address.
For multi-network card server, which kind of physical topology access network it adopts very big for the availability and the performance impact of network.At present, a kind of multi-network card server cut-in method comparatively commonly used is on multi-network card server, to use adaptability load-sharing mode (ALB, Adaptive Load Balance) to come access network.Fig. 1 is the topological sketch map when multi-network card server adopts ALB pattern access network in the prior art.Referring to Fig. 1; When adopting the ALB pattern, the mode of multi-network card server access network is: for all network interface cards in the multi-network card server, set out main with network interface card and subsequent use network interface card; And the operating state that will lead with network interface card and subsequent use network interface card all is set to state of activation (Active); Main with network interface card and the shared IP address of subsequent use network interface card, and all subsequent use network interface cards all use main medium access control (MAC) address with network interface card; All network interface cards on the multi-network card server all are connected on the switch in the network.
Referring to Fig. 1, when adopting the ALB pattern, thereby the professional process of multi-network card server access network realization comprises: at down direction, switch sends to the main network interface card of using according to the main MAC Address of learning with network interface card with downlink service data; At up direction, multi-network card server utilizes all network interface cards to carry out the load balancing of uplink traffic, that is to say, from the flow that multi-network card server sends, selects the particular port of a network interface card to transmit according to the traffic load situation on all network interface card ports.
Can find out by above description, when multi-network card server adopts the ALB pattern,, therefore, can satisfy the flow load balance of client's universal demand, have applied environment widely because it can realize the load balancing of up direction flow.But, when multi-network card server adopts the ALB pattern, because its all network interface cards must be connected on the switch in the network; Therefore, referring to Fig. 2, this switch then becomes single point failure point; That is to say, if this switch breaks down or the network equipment at this switch and the upper reaches between upstream Interface be connected and break down, so; The capital causes the service disconnection on the multi-network card server, the reliability when greatly reducing the multi-network card server access network.
Summary of the invention
In view of this, main purpose of the present invention is to provide a kind of cut-in method and system of multi-network card server, so that the reliability when improving the multi-network card server access network.
In order to achieve the above object, technical scheme of the present invention is achieved in that
A kind of cut-in method of multi-network card server is characterized in that, this method comprises:
A plurality of three-tier switch are formed the VRRP backup group, and the route priority of host exchange is higher than the route priority from switch in the configuration VRRP backup group;
Master in the multi-network card server is connected to the host exchange in the VRRP backup group with network interface card, subsequent use network interface card be connected in the VRRP backup group from switch; With host exchange in the VRRP backup group with link to each other from switch; Said multi-network card server is for using the multi-network card server of ALB pattern;
The host exchange downlink data receiving that has higher routing priority in the VRRP backup group utilizes the main address information of learning with network interface card that the downlink data that receives is sent to the major network card in the multi-network card server;
The virtual address information of the VRRP backup group that the network interface card utilization of confirming according to load-balancing algorithm in the multi-network card server obtains is sent the host exchange in upstream data to the VRRP backup group, and is transmitted by this host exchange.
A kind of system that realizes that multi-network card server inserts; It is characterized in that; This system comprises: by VRRP backup group and multi-network card server that a plurality of three-tier switch are formed, wherein, the route priority of host exchange is higher than the route priority from switch in the VRRP backup group; Master in the multi-network card server is connected to the host exchange in the VRRP backup group with network interface card, subsequent use network interface card be connected in the VRRP backup group from switch; In the said VRRP backup group host exchange with link to each other from switch; Described multi-network card server is for using the multi-network card server of ALB pattern;
The host exchange downlink data receiving that has higher routing priority in the VRRP backup group utilizes the main address information of learning with network interface card that the downlink data that receives is sent to the major network card in the multi-network card server, and transmits the upstream data that receives;
Any network interface card in the multi-network card server when sending upstream data according to load-balancing algorithm is definite by self, utilizes the virtual address information of the VRRP backup group that obtains to send the host exchange in upstream data to the VRRP backup group.
This shows; In the present invention, because the VRRP backup group that a plurality of three-tier switch is formed shows as a virtual switch for multi-network card server, like this; Requirement when all network interface cards of multi-network card server are connected to this virtual switch and have satisfied then that multi-network card server adopts the ALB pattern; In addition, owing in fact comprise a plurality of three-tier switch in the VRRP backup group that multi-network card server connects, can back up each other; Therefore, solved that the switch Single Point of Faliure causes the low shortcoming of service reliability in the prior art.
Further; The present invention is directed to the miscellaneous service scene, as the host exchange in the VRRP backup group, from switch or the connection between it take place unusual, main take place with network interface card or subsequent use network interface card unusual, situation such as the up link of host exchange is unusual; All provide the multi-network card server access network of ALB pattern to realize professional solution respectively; Therefore, further improve the reliability and the flexibility of business realizing, greatly strengthened practicality of the present invention.
Description of drawings
Fig. 1 is the network topology sketch map when multi-network card server adopts ALB pattern access network in the prior art.
Fig. 2 is the sketch map of switch Single Point of Faliure when multi-network card server adopts ALB pattern access network in the prior art.
Fig. 3 is a kind of network topology sketch map during the multi-network card server access network in one embodiment of the invention.
Fig. 4 works as multi-network card server, VRRP backup group and connects in one embodiment of the invention all to be in downlink traffic transfer process figure just often.
Fig. 5 works as multi-network card server, VRRP backup group and connects in one embodiment of the invention all to be in uplink traffic transfer process figure just often.
Fig. 6 is the network topology sketch map when the host exchange in the VRRP backup group is unusual in one embodiment of the invention.
Fig. 7 is when the network topology sketch map when unusual from switch in the VRRP backup group in one embodiment of the invention.
Fig. 8 the network topology sketch map when unusual occurs with network interface card as the master in the multi-network card server in one embodiment of the invention.
Fig. 9 the network topology sketch map when unusual occurs when the subsequent use network interface card in the multi-network card server in one embodiment of the invention.
Figure 10 is that the network topology sketch map when unusual appears in the up link that in one embodiment of the invention, connects when the host exchange in the VRRP backup group.
Figure 11 is that the downlink data transmission process chart when unusual appears in the up link that in one embodiment of the invention, connects when the host exchange in the VRRP backup group.
Figure 12 is that the transmitting uplink data process chart when unusual appears in the up link that in one embodiment of the invention, connects when the host exchange in the VRRP backup group.
Figure 13 is the network topology sketch map when host exchange in the VRRP backup group and being connected between switch are unusual in one embodiment of the invention.
Embodiment
For making the object of the invention, technical scheme and advantage clearer, the present invention is done to describe in detail further below in conjunction with accompanying drawing and specific embodiment.
When multi-network card server adopts the ALB pattern,, therefore, then can reduce professional reliability owing to switch becomes single point failure point because its all network interface cards must be connected on the switch in the network.For solving this shortcoming, the present invention proposes a kind of cut-in method of multi-network card server, this method comprises: a plurality of three-tier switch are formed Virtual Router Redundancy Protocol (VRRP) backup group; Master in the multi-network card server is connected to the host exchange in the VRRP backup group with network interface card, subsequent use network interface card be connected in the VRRP backup group from switch; Host exchange downlink data receiving in the VRRP backup group utilizes the main address information of learning with network interface card that the downlink data that receives is sent to the major network card in the multi-network card server; The network interface card of confirming according to load-balancing algorithm in the multi-network card server utilizes the virtual address information of VRRP backup group to send the host exchange in upstream data to the VRRP backup group, and is transmitted by this host exchange.
It is thus clear that; The inventive method utilizes that the VRRP backup group externally shows as a virtual unit in the VRRP technology, and in fact comprises the characteristics of a plurality of equipment in the backup group, and a plurality of three-tier switch are formed the VRRP backup groups; Because this VRRP backup group shows as a virtual switch for multi-network card server; Like this, the requirement when all network interface cards of multi-network card server are connected to this virtual switch and have satisfied then that multi-network card server adopts the ALB pattern, in addition; Owing in fact comprise a plurality of three-tier switch in the VRRP backup group that multi-network card server connects; Can back up each other, therefore, solve also that the switch Single Point of Faliure causes the low shortcoming of service reliability in the prior art.
Be example with the miscellaneous service scene respectively below, implementation procedure of the present invention at length is described.Wherein, the present invention not only adopts the backup of VRRP technical guarantee access switch, and; When guaranteeing that also multi-network card server adopts the ALB pattern; Satisfy the requirement of ALB pattern, that is, under normal circumstances; Downlink data only is transferred to the main network interface card of using, and upstream data is selected network interface card output and arrived the host exchange in the VRRP backup group according to load-balancing algorithm by all network interface cards.
Business scenario one, multi-network card server, VRRP backup group and all connections thereof all are in downlink traffic transmission just often.
Fig. 3 is a kind of network topology sketch map during the multi-network card server access network in one embodiment of the invention.Fig. 4 works as multi-network card server, VRRP backup group and connects all to be in downlink traffic transfer process figure just often in one embodiment of the invention.Referring to Fig. 3; For ease of describing; Main to comprise in the multi-network card server with network interface card and subsequent use network interface card totally two network interface cards, and the number of the three-tier switch in the VRRP backup group that connects of this server is 2 to be example, realizes that the complete procedure that downlink traffic transmits may further comprise the steps:
Step 401: in advance two three-tier switch are formed the VRRP backup groups, and will lead with network interface card and be connected to the host exchange in the VRRP backup group, with subsequent use network interface card be connected in the VRRP backup group from switch.
Through the processing of this step 401, make all network interface cards of multi-network card server be connected on the virtual switch of VRRP backup group formation, satisfied the requirement of a switch in needs access network when multi-network card server adopts the ALB pattern.
Step 402: in advance with the host exchange in the VRRP backup group with link to each other from switch.
Step 403: the route priority that host exchange in the VRRP backup group is set in advance is higher than the route priority from switch, and, this host exchange and externally issue its route precedence information from switch.
The processing of this step is can to confirm in the subsequent process downlink traffic to be transferred to the higher host exchange of route priority according to host exchange in the VRRP backup group and the route precedence information of issuing from switch for other equipment in the network that make.
Step 404: when the client-access multi-network card server, the equipment in the network is given the higher host exchange of VRRP backup group medium priority according to host exchange in the VRRP backup group with from the external route of issuing of switch with the downlink data transmission of this visit.
Because downlink data is the data of client-access multi-network card server, therefore, the purpose IP address in this downlink data is to use in the multi-network card server of ALB pattern main with network interface card and the shared IP address of subsequent use network interface card, is designated as IP1.
Step 405: after host exchange receives downlink data; Purpose IP address ip 1 according in this downlink data is searched, if can find corresponding address analysis protocol (ARP) list item and MAC address entries, and then direct execution in step 411; Otherwise, execution in step 406.
Step 406: host exchange sends the ARP request.
Here; When host exchange receives the downlink data between client and the multi-network card server first; Be that client is when issuing the request of data message of multi-network card server; Then can there be ARP list item corresponding and corresponding MAC address entries in the host exchange, need utilize the process of step 406 to step 410 to obtain main address information, the promptly corresponding ARP list item and the MAC address entries of correspondence with network interface card with IP 1.
Step 407: the ARP request of host exchange connects main port with network interface card from host exchange respectively sends with the port that is connected from switch, arrives main with network interface card with from switch respectively.
Here, referring to Fig. 3, the ARP of host exchange request transfers to the main network interface card of using through port a3 and port c, and transfers to from switch through port a2 and port b2.
Step 408: from switch the ARP request that receives is sent to subsequent use network interface card, subsequent use network interface card abandons received ARP request.
Here, referring to Fig. 3, the ARP request transfers to subsequent use network interface card through port b3 and port d.
In addition, it is in order to guarantee that switch can not set up ARP list item and the MAC address entries corresponding to subsequent use network interface card downlink data receiving that subsequent use network interface card abandons ARP request, thereby guarantees that downlink data can all be transferred to the main network interface card of using according to the requirement of ALB pattern.
Step 409: main receive the ARP request with network interface card after, return to host exchange and to carry this master with the IP address of network interface card and the arp response message of MAC Address.
Here, referring to Fig. 3, arp response message is back to host exchange through port c and port a3.
Step 410: after host exchange is received arp response message, utilize the IP address and the MAC Address that carry in this message to set up the ARP list item, and the port a3 that utilizes the MAC Address that carries in this message and receive this message sets up MAC address entries.
Here, the shared IP address of all network interface cards is IP1 in the multi-network card server, and shared MAC Address is MAC1, and so, the ARP list item of foundation is the corresponding relation of IP1 and MAC1, and the MAC address entries of foundation is the corresponding relation (MAC1-port a3) of MAC1 and port a3.
Utilize the process of above-mentioned steps 406 to step 410, host exchange has then been learnt main address information with network interface card.
Step 411: ARP list item and MAC address entries that the address information that the host exchange utilization is learnt is promptly set up directly are sent to the main network interface card of using with downlink data.
Here, host exchange is IP1 according to the purpose IP address of carrying in the downlink data, and ARP list item and the MAC address entries set up in the finding step 410 are finally determined port-for-port a3, therefore, through this port a3 downlink data is sent to the main network interface card of using.
So far, the present invention has realized that then multi-network card server, VRRP backup group and all connections thereof all are in downlink traffic transmission just often.
Business scenario two, multi-network card server, VRRP backup group and all connections thereof all are in uplink traffic transmission just often.
Fig. 5 works as multi-network card server, VRRP backup group and connects all to be in uplink traffic transfer process figure just often in one embodiment of the invention.Referring to Fig. 3 and Fig. 5; For ease of describing; Main to comprise in the multi-network card server with network interface card and subsequent use network interface card totally two network interface cards, and the number of the three-tier switch in the VRRP backup group that connects of this server is 2 to be example, realizes that the complete procedure that uplink traffic is transmitted may further comprise the steps:
Step 501～step 502 is all identical with all descriptions of step 401～402 steps.
Step 503: the master in the multi-network card server obtains the virtual address information (comprising virtual IP address and virtual MAC address) of VRRP backup group with network interface card.
Here, the multi-network card server mode of obtaining the virtual address information of VRRP backup group can realize according to client's demand flexibly.
Such as; Initiatively send the ARP request by the master in the multi-network card server with network interface card; After host exchange in the VRRP backup group receives this ARP request; Return the arp response message of the virtual address information (comprising virtual IP address and virtual MAC address) of carrying the VRRP backup group, the main virtual address information of from this arp response message, obtaining the VRRP backup group with network interface card.
For another example, also can be by leading with network interface card after receiving the ARP request that host exchange is sent, according to the virtual IP address of the VRRP backup group that carries in this ARP request and the virtual address information that the virtual MAC address obtains the VRRP backup group.Wherein, Main ARP request of receiving with network interface card can be that host exchange sends to main (such as can be referring to step 406 and the step 407 in the flow process shown in Figure 4) with network interface card in order to obtain the downlink transfer address in the downlink data transmission flow process; Perhaps; Also can be in order to realize transmitting uplink data, and by the host exchange special configuration in the VRRP backup group and the ARP request of sending.
And; After obtaining the virtual address information of VRRP backup group; Main form with network interface card preservation virtual address information is in multi-network card server, to set up the ARP list item, and the ARP list item of foundation is the virtual IP address of VRRP backup group and the corresponding relation (VIP-VMAC) of virtual MAC address.
Step 504: the master in the multi-network card server receives the request of data message from client with network interface card, confirms to send the network interface card and the port of upstream data according to load-balancing algorithm.
Here, when multi-network card server adopts the ALB pattern, need adopt load balancing for uplink traffic; Therefore, in this step, need to confirm this and carry out the network interface card and the corresponding port of response; Its concrete implementation procedure is identical with respective process of the prior art, no longer describes here.
Step 505: the network interface card of determining utilizes the virtual address information of VRRP backup group to send current host exchange in upstream data to the VRRP backup group.
Here; Referring to Fig. 3; When the network interface card of determining is main when being connected the port c of host exchange with network interface card and this master on network interface card, this is led with network interface card from port c transmission upstream data, and the source IP address of this upstream data and source MAC are respectively shared IP1 and the MAC1 of all network interface cards in this multi-network card server; Purpose IP address is the IP address of client, and target MAC (Media Access Control) address is the virtual address VMAC of VRRP backup group; This upstream data arrives host exchange through port a3.
Referring to Fig. 3; When the network interface card of determining be subsequent use network interface card with this subsequent use network interface card on when being connected the port d from switch; This subsequent use network interface card sends upstream data from port d; The source IP address of this upstream data and source MAC are respectively shared IP1 and the MAC1 of all network interface cards in this multi-network card server, and purpose IP address is the IP address of client, and target MAC (Media Access Control) address is the virtual address VMAC of VRRP backup group; This upstream data arrives from switch through port b3, gives host exchange from switch with transmitting uplink data.Wherein, need the following situation of explanation:
According to common processing; When host exchange is receiving behind the next upstream data of switch transmission with the port a2 that is connected from switch; Should learn MAC information; Set up the corresponding relation (MAC1-port a2) between source MAC MAC1 that carries in the upstream data and the port a2 that receives this upstream data, and, when from switch after the port b3 that is connected with subsequent use network interface card receives this upstream data; Should learn MAC information, set up the corresponding relation (MAC1-port b3) between source MAC MAC1 that carries in the upstream data and the port b3 that receives this upstream data.But; If allow host exchange to set up above-mentioned MAC information (MAC1-port a2) and set up above-mentioned MAC information (MAC1-port b3) from switch, so, in follow-up downlink data transmission process; Host exchange then can find (MAC1-port a2); Then downlink data transmission is given from switch, then can find (MAC1-port b3) from switch, thereby the downlink data mistake is dealt on the subsequent use network interface card; Can't satisfy downlink data in the ALB pattern and only be transferred to main requirement, thereby cause makeing mistakes with network interface card.For fear of the generation of above-mentioned mistake, preferably, in the present invention, forbid MAC address learning for the port arrangement that from switch, links to each other in advance with subsequent use network interface card, and, for forbidding MAC address learning with the port arrangement that links to each other from switch on the host exchange.Like this, in this step, from switch after port b3 receives upstream data; Then no longer carry out MAC address learning; Promptly no longer set up the corresponding relation between MAC1 and the port b3, after host exchange receives upstream data from port a2, also no longer carry out MAC address learning; Promptly no longer set up the corresponding relation between MAC1 and the port a2, thereby guaranteed business realization.
Step 506: host exchange sends the upstream data that receives at up direction.
So far, the present invention has realized that then multi-network card server, VRRP backup group and all connections thereof all are in uplink traffic transmission just often.
Processing when unusual appears in the host exchange in business scenario three, the VRRP backup group.
Fig. 6 is the network topology sketch map when the host exchange in the VRRP backup group is unusual in one embodiment of the invention.Referring to Fig. 6, host exchange in the VRRP backup group occurs unusual, in the time of can't operate as normal; This host exchange with from switch and main with breaking off respectively being connected then between the network interface card, at this moment, the processing that present embodiment provides comprises: the host exchange in the VRRP backup group unusually after; This host exchange switches to from switch, formerly switches to host exchange from switch, and the former master in the multi-network card server switches to subsequent use network interface card with network interface card; Former subsequent use network interface card switches to main use network interface card, and at this moment, multi-network card server only links to each other with the host exchange after network interface card and the switching through main after switching; Therefore; For downlink data transmission,, and directly be transferred to the main network interface card of using after the switching by the host exchange downlink data receiving after switching; For transmitting uplink data, with network interface card upstream data directly is transferred to the host exchange after the switching by the master after switching, the host exchange after this switches sends this upstream data on up direction.Behind the abnormal restoring of former host exchange, concrete downlink data transmission and transmitting uplink data are then respectively like above-mentioned Fig. 4 and flow process shown in Figure 5.
In business scenario four, the VRRP backup group from switch the processing when unusual appears.
Fig. 7 is when the network topology sketch map when unusual from switch in the VRRP backup group in one embodiment of the invention.Referring to Fig. 7, occurring from switch in the VRRP backup group is unusual, in the time of can't operate as normal; Should break off respectively from being connected then between switch and host exchange and the subsequent use network interface card, at this moment, the processing that present embodiment provides comprises: in the VRRP backup group from switch unusually after; Multi-network card server only links to each other with host exchange with network interface card through main, therefore, and for downlink data transmission; By the host exchange downlink data receiving, and directly be transferred to the main network interface card of using; For transmitting uplink data, with network interface card upstream data directly is transferred to host exchange by main, this host exchange sends this upstream data on up direction.When behind the abnormal restoring of switch, concrete downlink data transmission and transmitting uplink data are then respectively like above-mentioned Fig. 4 and flow process shown in Figure 5.
Master in business scenario five, the multi-network card server processing when unusual occurs with network interface card.
Fig. 8 the network topology sketch map when unusual occurs with network interface card as the master in the multi-network card server in one embodiment of the invention.Referring to Fig. 8, when the master in the multi-network card server occurs unusually can't operate as normal the time with network interface card, this master breaks off with being connected then between network interface card and the host exchange; At this moment, the processing that present embodiment provides comprises: the master in the multi-network card server with network interface card unusual after, this master switches to subsequent use network interface card with network interface card; Former subsequent use network interface card switches to the main network interface card of use, multi-network card server through after switching main with network interface card with link to each other from switch, and pass through to be somebody's turn to do and link to each other with host exchange from switch; Therefore, for downlink data transmission, by the host exchange downlink data receiving; And be transferred to from switch, again by giving the main network interface card of using after switching with this downlink data transmission from switch; For transmitting uplink data, transmitting uplink data is given from switch with network interface card by the master after switching, by giving host exchange from switch with transmitting uplink data, this host exchange sends this upstream data on up direction again.Behind former main abnormal restoring with network interface card, concrete downlink data transmission and transmitting uplink data are then respectively like above-mentioned Fig. 4 and flow process shown in Figure 5.
Processing when unusual appears in the subsequent use network interface card in business scenario six, the multi-network card server.
Fig. 9 the network topology sketch map when unusual occurs when the subsequent use network interface card in the multi-network card server in one embodiment of the invention.Referring to Fig. 9, when the subsequent use network interface card in the multi-network card server occurs unusually can't operate as normal the time, this subsequent use network interface card with break off from being connected then between the switch; At this moment, the processing that present embodiment provides comprises: after the subsequent use network interface card in the multi-network card server was unusual, multi-network card server only linked to each other with host exchange with network interface card through main; Therefore; For downlink data transmission,, and directly be transferred to the main network interface card of using by the host exchange downlink data receiving; For transmitting uplink data, with network interface card upstream data directly is transferred to host exchange by main, this host exchange sends this upstream data on up direction.Behind the abnormal restoring of subsequent use network interface card, concrete downlink data transmission and transmitting uplink data are then respectively like above-mentioned Fig. 4 and flow process shown in Figure 5.
The processing of downlink data transmission when unusual appears in the up link that host exchange in business scenario seven, the VRRP backup group connects.
Figure 10 is that the network topology sketch map when unusual appears in the up link that in one embodiment of the invention, connects when the host exchange in the VRRP backup group.Figure 11 is that the downlink data transmission process chart when unusual appears in the up link that in one embodiment of the invention, connects when the host exchange in the VRRP backup group.Referring to Figure 10 and Figure 11, this handling process may further comprise the steps:
Host exchange in the step 1101:VRRP backup group detects the state of its uplink port; If uplink port aborted; Then reduce local VRRP priority; This host exchange in the VRRP backup group then switches to from switch, formerly then switches to host exchange from switch, and the activestandby state of each network interface card remains unchanged in the multi-network card server.
Step 1102: when the client-access multi-network card server, the equipment in the network is unusual according to above-mentioned up link, gives the host exchange (what link to each other with subsequent use network interface card is former in switch) after switching in the VRRP backup group with the downlink data transmission of this visit.
Because downlink data is the data of client-access multi-network card server, therefore, the purpose IP address in this downlink data is to use in the multi-network card server of ALB pattern main with network interface card and the shared IP address of subsequent use network interface card, is designated as IP1.
Step 1103: after the host exchange after the switching receives downlink data; Purpose IP address ip 1 according in this downlink data is searched, if can find corresponding ARP list item and MAC address entries, and then direct execution in step 1110; Otherwise, execution in step 1104.
Step 1104: the host exchange after the switching sends the ARP request.
Step 1105: the ARP request of the host exchange after the switching is sent with the port from switch (former host exchange) that is connected after switching from the port that connects subsequent use network interface card respectively; After arriving subsequent use network interface card respectively and switching from switch, subsequent use network interface card abandons received ARP request.
Here, referring to Fig. 3, the ARP of the host exchange after switching request transfers to subsequent use network interface card through port b3 and port d.And through port b2, a2 arrive after switching from switch.
Step 1106: from switch the ARP that receives request is sent to the main network interface card of using after the switching.
Here, referring to Fig. 3, ARP request transfers to subsequent use network interface card from switch through port a3 and port c after the switching.
In addition, to abandon ARP request be in order to guarantee that switch can not set up the ARP list item corresponding to subsequent use network interface card downlink data receiving to subsequent use network interface card in the step 1105, thereby guarantee that downlink data can all be transferred to the main network interface card of using according to the requirement of ALB pattern.
Step 1107: after the master received the ARP request with network interface card, returning from switch after switching carried this master with the IP address of network interface card and the arp response message of MAC Address.
Step 1108: after the switching from switch with the host exchange of this arp response transmission of messages after switch.
Step 1109: after the host exchange after the switching is received arp response message, utilize the IP address and the MAC Address that carry in this message to set up the ARP list item, and the port that utilizes the MAC Address that carries in this message and receive this message is set up MAC address entries.
Here; If the shared IP address of all network interface cards is IP1 in the multi-network card server, shared MAC Address is MAC1, so; The ARP list item of setting up is the corresponding relation of IP1 and MAC1, and the MAC address entries of foundation is the corresponding relation (MAC1-port b2) of MAC1 and port b2.
Utilize the process of above-mentioned steps 1104 to step 1109, the host exchange after the switching has then been learnt main address information with network interface card.
Step 1110: ARP list item and MAC address entries that the address information that the host exchange utilization after the switching is learnt is promptly set up, with downlink data through being sent to the main network interface card of using from switch after switching.
Here; Host exchange after the switching is IP1 according to the purpose IP address of carrying in the downlink data; Search ARP list item and MAC address entries, finally determine port-for-port b2, therefore; Through this port b2 with downlink data be sent to after the switching from switch, from switch downlink data is sent to the main network interface card of using after this switching.
So far, the present invention has realized that then the downlink data transmission process chart when unusual appears in up link that host exchange connects.
The processing of transmitting uplink data when unusual appears in the up link that host exchange in business scenario eight, the VRRP backup group connects.
Figure 12 is that the transmitting uplink data process chart when unusual appears in the up link that in one embodiment of the invention, connects when the host exchange in the VRRP backup group.Referring to Figure 10 and Figure 12, this handling process may further comprise the steps:
Step 1201～step 1202 is all identical with all descriptions of step 1101～step 1102.
Step 1203～step 1204 is identical with the implementation procedure of step 503～step 504.
Step 1205: the network interface card of determining utilizes the virtual address information of VRRP backup group to send the host exchange after switching in upstream data to the VRRP backup group.
Here; Referring to Fig. 3; When the network interface card of determining is main when being connected the port c of host exchange with network interface card and this master on network interface card, this is led with network interface card from port c transmission upstream data, and the source IP address of this upstream data and source MAC are respectively shared IP1 and the MAC1 of all network interface cards in this multi-network card server; Purpose IP address is the IP address of client, and target MAC (Media Access Control) address is the virtual address VMAC of VRRP backup group; This upstream data after port a3 arrive to switch from switch, after the switching from switch with the host exchange of this transmitting uplink data after switch.
Referring to Fig. 3; When the network interface card of determining be subsequent use network interface card with this subsequent use network interface card on when being connected the port d from switch; This subsequent use network interface card sends upstream data from port d; The source IP address of this upstream data and source MAC are respectively shared IP1 and the MAC1 of all network interface cards in this multi-network card server, and purpose IP address is the IP address of client, and target MAC (Media Access Control) address is the virtual address VMAC of VRRP backup group; This upstream data arrives the host exchange after switching through port b3.
Because the description according to step 505 place in the above-mentioned flow process shown in Figure 5 can obtain; Former port from the subsequent use network interface card of switch connection is pre-configured forbids MAC address learning, therefore, and in this step; Though port b3 goes up and has received upstream data, can not learn MAC Address.
Step 1206: the host exchange after the switching sends the upstream data that receives at up direction.
So far, the present invention has realized that then the transmitting uplink data process chart when unusual appears in up link that host exchange connects.
Processing when host exchange is unusual with being connected between switch in business scenario nine, the VRRP backup group.
Figure 13 is the network topology sketch map when host exchange in the VRRP backup group and being connected between switch are unusual in one embodiment of the invention.Referring to Figure 13; This handling process comprises: when host exchange in the VRRP backup group and being connected between switch were unusual, this connected disconnection, therefore; The VRRP backup group in fact no longer exists; Then former host exchange is still remained host exchange, and also upgrade to host exchange with former from switch, promptly main what be connected separately with network interface card and subsequent use network interface card all is host exchange; For downlink data transmission, the host exchange downlink data receiving higher by priority also directly is transferred to the main network interface card of using; For transmitting uplink data; Confirm to send the network interface card and the port of upstream data according to load-balancing algorithm; If what determine is the main network interface card of using; Then mainly utilize the virtual address information of VRRP backup group to send upstream data to direct-connected host exchange with network interface card, this host exchange sends this upstream data on up direction, if what determine is subsequent use network interface card; Then subsequent use network interface card utilizes the host exchange (former in switch) after the virtual address information of VRRP backup group is sent upstream data to direct-connected upgrading, and the host exchange after this upgrading sends this upstream data on up direction.
Processing when so far, the present invention has realized then that host exchange in the VRRP backup group and being connected between switch are unusual.
In addition, the invention allows for a kind of system that realizes that multi-network card server inserts.This system comprises: VRRP backup group and the multi-network card server be made up of a plurality of three-tier switch; Wherein, The route priority of host exchange is higher than the route priority from switch in the VRRP backup group; Master in the multi-network card server is connected to the host exchange in the VRRP backup group with network interface card, subsequent use network interface card be connected in the VRRP backup group from switch;
The host exchange downlink data receiving that has higher routing priority in the VRRP backup group utilizes the main address information of learning with network interface card that the downlink data that receives is sent to the major network card in the multi-network card server, and transmits the upstream data that receives;
Any network interface card in the multi-network card server when sending upstream data according to load-balancing algorithm is definite by self, utilizes the virtual address information of the VRRP backup group that obtains to send the host exchange in upstream data to the VRRP backup group.
For system proposed by the invention, a kind of possible implementation is a structure as shown in Figure 3.
For system of the present invention, realize that host exchange comprises downlink data transmission to main a kind of feasible implementation with network interface card:
Said host exchange with link to each other from switch;
Said host exchange; Send the ARP request from connecting main port with being connected from the port of switch with network interface card; After receiving arp response message; Utilize to lead in this message and set up the ARP list item with the IP address of network interface card and the corresponding relation of MAC Address; And utilize main MAC Address and the corresponding relation that receives the port of arp response message with network interface card to set up MAC address entries, and mainly search ARP list item and MAC address entries successively according to what carry in the downlink data with the IP address of network interface card, send downlink data from the port that finds and use network interface card to the master;
Said from switch, the ARP request that receives is sent to the subsequent use network interface card that it links to each other;
Said subsequent use network interface card abandons received ARP request;
Saidly main use network interface card, receive after ARP asks, return to host exchange and carry this master with the IP address of network interface card and the arp response message of MAC Address.
For system of the present invention, realize comprising to a kind of feasible implementation of host exchange transmitting uplink data:
Said host exchange with link to each other from switch, connect on the host exchange from the port of switch and be configured to forbid MAC address learning, the port that connects subsequent use network interface card from switch is configured to forbid MAC address learning;
Said subsequent use network interface card is when sending upstream data according to load-balancing algorithm is definite by self, through with the port that links to each other from switch upstream data being sent to from switch;
Said from switch, further with port that subsequent use network interface card is connected on when receiving upstream data, no longer carry out MAC address learning, directly this upstream data is sent to host exchange;
Said host exchange is further receiving upstream data with the port that is connected from switch, no longer carries out MAC address learning.
Further the reply miscellaneous service is unusual in the system that the present invention proposes, and, when various abnormal conditions take place, can realize that still the multi-network card server access network carries out service communication that is, can comprise such as concrete implementation:
When the host exchange in the VRRP backup group was unusual, this host exchange switched to from switch, formerly switched to host exchange from switch, and the master in the multi-network card server switches to subsequent use network interface card with network interface card, and former subsequent use network interface card switches to the main network interface card of using; By the host exchange downlink data receiving after switching, and directly be transferred to the main network interface card of using after the switching; With network interface card upstream data directly is transferred to the host exchange after the switching by the master after switching, the host exchange after this switching sends upstream data on up direction;
And/or,
When unusual,, and directly being transferred to the main network interface card of using from switch in the VRRP backup group by the host exchange downlink data receiving; With network interface card upstream data directly is transferred to host exchange by main, this host exchange sends this upstream data on up direction;
And/or,
When the master in the multi-network card server is unusual with network interface card; This master switches to subsequent use network interface card with network interface card, and former subsequent use network interface card switches to the master and uses network interface card, by the host exchange downlink data receiving; And be transferred to from switch, again by giving the main network interface card of using after switching with this downlink data transmission from switch; Transmitting uplink data is given from switch with network interface card by the master after switching, by giving host exchange from switch with transmitting uplink data, this host exchange sends this upstream data on up direction again;
And/or,
When the subsequent use network interface card in the multi-network card server is unusual,, and directly be transferred to the main network interface card of using by the host exchange downlink data receiving; With network interface card upstream data directly is transferred to host exchange by main, this host exchange sends this upstream data on up direction.
The system that the present invention proposes is the following service exception of reply further, comprising:
Host exchange in the said VRRP backup group; Further after detecting the uplink port aborted; Switch to from switch, the downlink data that the host exchange after switching is sent directly is sent to the main network interface card of using, and; After receiving the main upstream data of sending with network interface card, with upstream data be sent to after the switching from switch;
Saidly further switch to host exchange from switch; Host exchange downlink data receiving after this switches; The forwarding address information that utilization is learnt with this downlink data send to the main switching that directly links to each other with network interface card after from switch; And, the upstream data of sending from switch after switching is transmitted.
The system that the present invention proposes is the following service exception of reply further, comprising:
Said host exchange further detecting when unusual with being connected between switch, remains host exchange, and downlink data receiving also directly is transferred to and mainly use network interface card, receives when leading the upstream data of sending with network interface card, on up direction, sends;
Said from switch, further, upgrade to host exchange detecting when unusual with being connected between switch, when receiving the upstream data that subsequent use network interface card sends, on up direction, send.
In a word, the above is merely preferred embodiment of the present invention, is not to be used to limit protection scope of the present invention.All within spirit of the present invention and principle, any modification of being done, be equal to replacement, improvement etc., all should be included within protection scope of the present invention.
Claims (18)
1. the cut-in method of a multi-network card server is characterized in that, this method comprises:
A plurality of three-tier switch are formed the VRRP backup group, and the route priority of host exchange is higher than the route priority from switch in the configuration VRRP backup group;
Master in the multi-network card server is connected to the host exchange in the VRRP backup group with network interface card, subsequent use network interface card be connected in the VRRP backup group from switch; With host exchange in the VRRP backup group with link to each other from switch; Described multi-network card server is for using the multi-network card server of adaptability load balancing ALB pattern;
The host exchange downlink data receiving that has higher routing priority in the VRRP backup group utilizes the main address information of learning with network interface card that the downlink data that receives is sent to the major network card in the multi-network card server;
The virtual address information of the VRRP backup group that the network interface card utilization of confirming according to load-balancing algorithm in the multi-network card server obtains is sent the host exchange in upstream data to the VRRP backup group, and is transmitted by this host exchange.
2. method according to claim 1 is characterized in that, this method further comprises: connect the port blocking MAC address learning from switch on the configuration host exchange, and connect the port blocking MAC address learning of subsequent use network interface card from switch.
3. method according to claim 1 is characterized in that, this method further comprises: the host exchange in the said VRRP backup group and from switch respectively network the issue its route precedence information;
The step that has the host exchange downlink data receiving of higher priority in the said VRRP backup group comprises: the equipment in the network is given downlink data transmission the host exchange that has higher routing priority in the VRRP backup group according to host exchange with from the route precedence information of switch issue.
4. method according to claim 1 is characterized in that,
Host exchange study master comprises with the process of the address information of network interface card in the VRRP backup group: host exchange sends the ARP request from connecting main port with network interface card with being connected from the port of switch; From switch the ARP request that receives is sent to the subsequent use network interface card that it links to each other, subsequent use network interface card abandons received ARP request; Main receive the ARP request with network interface card after, return to host exchange and to carry this master with the IP address of network interface card and the arp response message of MAC Address; The host exchange utilization is main sets up the ARP list item with the IP address of network interface card and the corresponding relation of MAC Address, and utilizes main MAC Address and the corresponding relation that receives the port of arp response message with network interface card to set up MAC address entries;
The said step that downlink data is sent to the major network card comprises: mainly search ARP list item and MAC address entries successively with the IP address of network interface card according to what carry in the downlink data, from the port transmission downlink data that finds.
5. method according to claim 1; It is characterized in that; The process that obtains the virtual address information of VRRP backup group comprises: the master sends the ARP request with network interface card, and host exchange returns arp response message after receiving this ARP request; Carry the virtual address information that comprises virtual IP address and virtual MAC address of VRRP backup group in this message, the main virtual address information of from this arp response message, obtaining the VRRP backup group with network interface card.
6. method according to claim 1; It is characterized in that; When the network interface card of confirming according to load-balancing algorithm is main when use network interface card, the step of the host exchange in said transmission upstream data to the VRRP backup group comprises: lead and through the port that links to each other with host exchange upstream data directly is sent to the host exchange in the VRRP backup group with network interface card;
Or,
This method further comprises: connect the port blocking MAC address learning from switch on the configuration host exchange, and connect the port blocking MAC address learning of subsequent use network interface card from switch; And,
When the network interface card of confirming according to load-balancing algorithm is subsequent use network interface card; The step of the host exchange in said transmission upstream data to the VRRP backup group comprises: subsequent use network interface card is through being sent to upstream data from switch with the port that links to each other from switch; From switch when the port that is connected with subsequent use network interface card receives upstream data; No longer carry out MAC address learning; Directly this upstream data is sent to host exchange, host exchange no longer carries out MAC address learning receiving upstream data with the port that is connected from switch.
7. according to any described method in the claim 1 to 6, it is characterized in that this method further comprises:
When the host exchange in the VRRP backup group was unusual, this host exchange switched to from switch, formerly switched to host exchange from switch, and the master in the multi-network card server switches to subsequent use network interface card with network interface card, and former subsequent use network interface card switches to the main network interface card of using; By the host exchange downlink data receiving after switching, and directly be transferred to the main network interface card of using after the switching; With network interface card upstream data directly is transferred to the host exchange after the switching by the master after switching, the host exchange after this switching sends upstream data on up direction;
Or,
When unusual,, and directly being transferred to the main network interface card of using from switch in the VRRP backup group by the host exchange downlink data receiving; With network interface card upstream data directly is transferred to host exchange by main, this host exchange sends this upstream data on up direction;
Or,
When the master in the multi-network card server is unusual with network interface card; This master switches to subsequent use network interface card with network interface card, and former subsequent use network interface card switches to the master and uses network interface card, by the host exchange downlink data receiving; And be transferred to from switch, again by giving the main network interface card of using after switching with this downlink data transmission from switch; Transmitting uplink data is given from switch with network interface card by the master after switching, by giving host exchange from switch with transmitting uplink data, this host exchange sends this upstream data on up direction again;
Or,
When the subsequent use network interface card in the multi-network card server is unusual,, and directly be transferred to the main network interface card of using by the host exchange downlink data receiving; With network interface card upstream data directly is transferred to host exchange by main, this host exchange sends this upstream data on up direction.
8. according to any described method in the claim 1 to 6; It is characterized in that; This method further comprises: the host exchange in the VRRP backup group detects the uplink port aborted; Switch to from switch, formerly switch to host exchange from switch, the activestandby state of each network interface card remains unchanged in the multi-network card server; And,
By the host exchange downlink data receiving after switching, utilize the forwarding address information learnt with this downlink data send to the main switching that directly links to each other with network interface card after from switch, from switch downlink data directly being sent to lead after this switching used network interface card; And/or the virtual address information of the VRRP backup group that the network interface card utilization of confirming according to load-balancing algorithm in the multi-network card server obtains is sent the host exchange after switching in upstream data to the VRRP backup group, and is transmitted by the host exchange after this switching.
9. method according to claim 8; It is characterized in that; The process of the host exchange study forwarding address information after the switching comprises: the host exchange after the switching sends the ARP request from the port that connects subsequent use network interface card with the port from switch that is connected after switching, and subsequent use network interface card abandons received ARP request; From switch the ARP that receives request is sent to main network interface card of using that it links to each other after the switching; Main carry this master with the IP address of network interface card and the arp response message of MAC Address, this arp response message is sent to the host exchange after the switching from switch after the switching with network interface card returning after switching from switch; Host exchange utilization after the switching is main sets up the ARP list item with the IP address of network interface card and the corresponding relation of MAC Address, and utilizes main MAC Address and the corresponding relation that receives the port of arp response message with network interface card to set up MAC address entries;
The said step from switch that downlink data is sent to after the switching comprises: mainly search ARP list item and the MAC address entries of being set up successively with the IP address of network interface card according to what carry the downlink data, from the port transmission downlink data that finds.
10. method according to claim 8; It is characterized in that; When the network interface card of confirming according to load-balancing algorithm was subsequent use network interface card, the step of the host exchange after switching in said transmission upstream data to the VRRP backup group comprised: subsequent use network interface card through with switch after the port that links to each other of host exchange upstream data directly is sent to the host exchange after the switching in the VRRP backup group;
Or,
When the network interface card of confirming according to load-balancing algorithm is main when using network interface card; The step of the host exchange after switching in said transmission upstream data to the VRRP backup group comprises: main with network interface card with upstream data be sent to after the switching from switch, this upstream data is sent to the host exchange after the switching from switch after the switching.
11. according to any described method in the claim 1 to 6; It is characterized in that; This method further comprises: when host exchange in the VRRP backup group and being connected between switch were unusual, former host exchange still remained host exchange, with former be host exchange from switch upgrades; The host exchange downlink data receiving higher by priority also directly is transferred to the main network interface card of using; When using the master to send upstream data according to load-balancing algorithm with network interface card; The main virtual address information transmission upstream data that utilizes the VRRP backup group with network interface card is to direct-connected host exchange; This host exchange sends this upstream data on up direction; When using subsequent use network interface card to send upstream data according to load-balancing algorithm; Subsequent use network interface card utilizes the host exchange after the virtual address information of VRRP backup group is sent upstream data to direct-connected upgrading, and the host exchange after this upgrading sends this upstream data on up direction.
12. system that realizes that multi-network card server inserts; It is characterized in that; This system comprises: by VRRP backup group and multi-network card server that a plurality of three-tier switch are formed, wherein, the route priority of host exchange is higher than the route priority from switch in the VRRP backup group; Master in the multi-network card server is connected to the host exchange in the VRRP backup group with network interface card, subsequent use network interface card be connected in the VRRP backup group from switch; In the said VRRP backup group host exchange with link to each other from switch; Described multi-network card server is for using the multi-network card server of adaptability load balancing ALB pattern;
The host exchange downlink data receiving that has higher routing priority in the VRRP backup group utilizes the main address information of learning with network interface card that the downlink data that receives is sent to the major network card in the multi-network card server, and transmits the upstream data that receives;
Any network interface card in the multi-network card server when sending upstream data according to load-balancing algorithm is definite by self, utilizes the virtual address information of the VRRP backup group that obtains to send the host exchange in upstream data to the VRRP backup group.
13. system according to claim 12; It is characterized in that; Host exchange in the said VRRP backup group is with adjacent from switch; And connect this port on this host exchange and be configured to forbid MAC address learning, should be configured to forbid MAC address learning from the port of the subsequent use network interface card of switch connection from switch.
14. system according to claim 12 is characterized in that,
Said host exchange; Send the ARP request from connecting main port with being connected from the port of switch with network interface card; After receiving arp response message; Utilize to lead in this message and set up the ARP list item with the IP address of network interface card and the corresponding relation of MAC Address; And utilize main MAC Address and the corresponding relation that receives the port of arp response message with network interface card to set up MAC address entries, and mainly search ARP list item and MAC address entries successively according to what carry in the downlink data with the IP address of network interface card, send downlink data from the port that finds and use network interface card to the master;
Said from switch, the ARP request that receives is sent to the subsequent use network interface card that it links to each other;
Said subsequent use network interface card abandons received ARP request;
Saidly main use network interface card, receive after ARP asks, return to host exchange and carry this master with the IP address of network interface card and the arp response message of MAC Address.
15. system according to claim 12 is characterized in that, connects from the port of switch on the host exchange to be configured to forbid MAC address learning, the port that connects subsequent use network interface card from switch is configured to forbid MAC address learning;
Said subsequent use network interface card is when sending upstream data according to load-balancing algorithm is definite by self, through with the port that links to each other from switch upstream data being sent to from switch;
Said from switch, further with port that subsequent use network interface card is connected on when receiving upstream data, no longer carry out MAC address learning, directly this upstream data is sent to host exchange;
Said host exchange is further receiving upstream data with the port that is connected from switch, no longer carries out MAC address learning.
16. system according to claim 12 is characterized in that,
When the host exchange in the VRRP backup group was unusual, this host exchange switched to from switch, formerly switched to host exchange from switch, and the master in the multi-network card server switches to subsequent use network interface card with network interface card, and former subsequent use network interface card switches to the main network interface card of using; By the host exchange downlink data receiving after switching, and directly be transferred to the main network interface card of using after the switching; With network interface card upstream data directly is transferred to the host exchange after the switching by the master after switching, the host exchange after this switching sends upstream data on up direction;
Or,
When unusual,, and directly being transferred to the main network interface card of using from switch in the VRRP backup group by the host exchange downlink data receiving; With network interface card upstream data directly is transferred to host exchange by main, this host exchange sends this upstream data on up direction;
Or,
When the master in the multi-network card server is unusual with network interface card; This master switches to subsequent use network interface card with network interface card, and former subsequent use network interface card switches to the master and uses network interface card, by the host exchange downlink data receiving; And be transferred to from switch, again by giving the main network interface card of using after switching with this downlink data transmission from switch; Transmitting uplink data is given from switch with network interface card by the master after switching, by giving host exchange from switch with transmitting uplink data, this host exchange sends this upstream data on up direction again;
Or,
When the subsequent use network interface card in the multi-network card server is unusual,, and directly be transferred to the main network interface card of using by the host exchange downlink data receiving; With network interface card upstream data directly is transferred to host exchange by main, this host exchange sends this upstream data on up direction.
17. according to any described system in the claim 12 to 16, it is characterized in that,
Host exchange in the said VRRP backup group; Further after detecting the uplink port aborted; Switch to from switch, the downlink data that the host exchange after switching is sent directly is sent to the main network interface card of using, and; After receiving the main upstream data of sending with network interface card, upstream data is sent to the host exchange after the switching;
Saidly further switch to host exchange from switch; Host exchange downlink data receiving after this switches; The forwarding address information that utilization is learnt with this downlink data send to the main switching that directly links to each other with network interface card after from switch; And, the upstream data of sending from switch after switching is sent at up direction.
18. according to any described system in the claim 12 to 16, it is characterized in that,
Said host exchange further detecting when unusual with being connected between switch, remains host exchange, and downlink data receiving also directly is transferred to and mainly use network interface card, receives when leading the upstream data of sending with network interface card, on up direction, sends;
Said from switch, further, upgrade to host exchange detecting when unusual with being connected between switch, when receiving the upstream data that subsequent use network interface card sends, on up direction, send.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009100766005A CN101488918B (en) | 2009-01-09 | 2009-01-09 | Multi-network card server access method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009100766005A CN101488918B (en) | 2009-01-09 | 2009-01-09 | Multi-network card server access method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101488918A CN101488918A (en) | 2009-07-22 |
| CN101488918B true CN101488918B (en) | 2012-02-08 |
Family
ID=40891600
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2009100766005A Active CN101488918B (en) | 2009-01-09 | 2009-01-09 | Multi-network card server access method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101488918B (en) |
Families Citing this family (31)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102694714B (en) * | 2011-03-21 | 2015-01-21 | 华为技术有限公司 | Method, device, system and a non-forwarder non-forwarding bridge for message transmission in TRILL network |
| CN102123105B (en) * | 2011-04-20 | 2013-12-11 | 杭州华三通信技术有限公司 | Method and equipment for switching between standard VRRP (Virtual Router Redundancy Protocol) and load balancing VRRP |
| CN102237929B (en) * | 2011-07-15 | 2015-02-25 | 北京星网锐捷网络技术有限公司 | Equipment switching method, device and system based on fiber communication network |
| CN102413047A (en) * | 2011-11-24 | 2012-04-11 | 华为技术有限公司 | Three-layer network device and route distribution method of same |
| CN102497430B (en) * | 2011-12-13 | 2014-12-03 | 曙光信息产业(北京)有限公司 | System and method for implementing splitting equipment |
| CN102868630B (en) * | 2012-09-07 | 2015-08-05 | 北京邮电大学 | A kind of parallel adaptive data transmission method for uplink based on many IP address and device |
| CN103795603B (en) | 2012-11-01 | 2017-08-11 | 新华三技术有限公司 | A kind of edge based on many network interface cards virtual bridged implementation method and equipment |
| CN103036780A (en) * | 2012-12-14 | 2013-04-10 | 滁州安瑞电力自动化有限公司 | Intelligent hot standby redundant double network cards and hot standby redundant method |
| CN103152256B (en) * | 2013-02-22 | 2017-05-03 | 浪潮电子信息产业股份有限公司 | Virtual routing network design method based on cloud computing data center |
| CN104978174B (en) * | 2014-04-08 | 2018-04-27 | 华为技术有限公司 | Switching method and system in multiple operating system between network interface card |
| CN103986665A (en) * | 2014-05-12 | 2014-08-13 | 浪潮电子信息产业股份有限公司 | Method for balancing network load in cloud sea OS |
| CN104125303B (en) * | 2014-08-12 | 2018-09-04 | 北京奇虎科技有限公司 | Reading and writing data requesting method, client and system |
| CN105791176A (en) * | 2014-12-23 | 2016-07-20 | 中兴通讯股份有限公司 | Method and device for determining IT (Information Technology) equipment port |
| CN104702518B (en) * | 2015-03-05 | 2017-12-19 | 安徽清新互联信息科技有限公司 | A kind of more card router apparatus and its data transmission method based on wireless network |
| US10298435B2 (en) * | 2015-06-08 | 2019-05-21 | Quanta Computer Inc. | Server link state detection and notification |
| CN106559246B (en) | 2015-09-30 | 2020-01-10 | 新华三技术有限公司 | Cluster implementation method and server |
| CN106100987B (en) * | 2016-08-08 | 2019-06-07 | 杭州迪普科技股份有限公司 | A kind of active router based on VRRP agreement switches the method and device of secondary router |
| CN106452913B (en) * | 2016-11-22 | 2020-09-01 | 卡斯柯信号有限公司 | Method for multiplexing and fast switching terminal network equipment |
| CN107342894A (en) * | 2017-06-22 | 2017-11-10 | 光大环保能源(常州)有限公司 | A kind of communication system and its control system and implementation method for waste incinerator |
| CN108055338B (en) * | 2017-12-22 | 2021-04-09 | 创新科技术有限公司 | ISCSI access load balancing method |
| CN108429675B (en) * | 2018-03-30 | 2021-10-12 | 新华三技术有限公司 | VRRP (virtual router redundancy protocol) linkage switching method and device and electronic equipment |
| CN108718398A (en) * | 2018-05-25 | 2018-10-30 | 苏州科达科技股份有限公司 | Code stream transmission method, device and the conference facility of video conferencing system |
| CN109347656B (en) * | 2018-09-11 | 2022-02-18 | 郑州云海信息技术有限公司 | Method and system for configuring virtual host uplink network card |
| CN109347768A (en) * | 2018-10-31 | 2019-02-15 | 四川中企互信信息技术有限公司 | A kind of method for switching network applied to double netcard |
| CN110198337B (en) * | 2019-03-04 | 2021-10-08 | 腾讯科技（深圳）有限公司 | Network load balancing method and device, computer readable medium and electronic equipment |
| CN111030926B (en) * | 2019-12-20 | 2021-07-27 | 苏州浪潮智能科技有限公司 | Method and device for improving high availability of network |
| CN111585800B (en) * | 2020-04-29 | 2022-06-28 | 平安科技（深圳）有限公司 | Virtual private cloud-based network node configuration method, device and medium |
| CN112910988A (en) * | 2021-01-28 | 2021-06-04 | 网宿科技股份有限公司 | Resource acquisition method and resource scheduling device |
| US20230077521A1 (en) * | 2021-09-14 | 2023-03-16 | Dell Products L.P. | Managing failover between information handling systems |
| CN115277726A (en) * | 2022-05-30 | 2022-11-01 | 浪潮软件集团有限公司 | Cluster data transmission method and system for dual networks |
| CN115514767B (en) * | 2022-09-27 | 2024-09-17 | 上汽通用五菱汽车股份有限公司 | Switching method of data transmission, terminal equipment and computer readable storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6282201B1 (en) * | 1997-11-20 | 2001-08-28 | Cisco Technology, Inc. | Method for configuring distributed internet protocol gateways with LAN emulation |
| CN1359213A (en) * | 2000-05-24 | 2002-07-17 | 阿尔卡塔尔互联网运行公司 | Method and equipment for supplying multiple protocol redundant router protocol support |
| CN1925496A (en) * | 2006-09-15 | 2007-03-07 | 杭州华为三康技术有限公司 | System and method for load sharing of network layer with multiple network interface cards terminal equipment |
| CN101051951A (en) * | 2006-05-29 | 2007-10-10 | 华为技术有限公司 | Method and device for securing server connection reliability |
-
2009
- 2009-01-09 CN CN2009100766005A patent/CN101488918B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6282201B1 (en) * | 1997-11-20 | 2001-08-28 | Cisco Technology, Inc. | Method for configuring distributed internet protocol gateways with LAN emulation |
| CN1359213A (en) * | 2000-05-24 | 2002-07-17 | 阿尔卡塔尔互联网运行公司 | Method and equipment for supplying multiple protocol redundant router protocol support |
| CN101051951A (en) * | 2006-05-29 | 2007-10-10 | 华为技术有限公司 | Method and device for securing server connection reliability |
| CN1925496A (en) * | 2006-09-15 | 2007-03-07 | 杭州华为三康技术有限公司 | System and method for load sharing of network layer with multiple network interface cards terminal equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101488918A (en) | 2009-07-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101488918B (en) | Multi-network card server access method and system | |
| EP3522462B1 (en) | Selecting an external link of a plurality of external links | |
| EP2109962B1 (en) | Triple-tier anycast addressing | |
| US8014409B1 (en) | Virtual router identifier that spans multiple interfaces in a routing device | |
| EP2640013B1 (en) | Method And Apparatus Providing Network Redundancy And High Availability To Remote Network Nodes | |
| US6151297A (en) | Method and system for link level server/switch trunking | |
| CN1754374B (en) | Network address translation with gateway load distribution | |
| US7185106B1 (en) | Providing services for multiple virtual private networks | |
| KR100310652B1 (en) | Method and System for Improving Communications in Data Communications Networks that provide Network Emulation | |
| US20040078772A1 (en) | Dynamic route exchange | |
| CN102714668A (en) | Distributed virtual fibre channel over Ethernet forwarder | |
| JPWO2011065268A1 (en) | Load distribution system, load distribution method, and program | |
| CN104754025A (en) | Programmable Distributed Networking | |
| US20100002714A1 (en) | PCI express network | |
| CN113259235B (en) | IPv 6-based dual-active route redundancy method and system | |
| GB2485026A (en) | Routed Split Multi-Link Trunking (RSMLT) resiliency for Wireless Local Area Network (WLAN) split-plane environments | |
| CN105282631A (en) | System and method for realizing electrical layer linear protection in POTN | |
| CN102291305A (en) | Method and device for implementing 6 to 4 relay routing, and message forwarding method | |
| CN106375355B (en) | Load balancing processing method and device | |
| CN101795238B (en) | Network load balancing networking method, network load balancing networking equipment and network load balancing networking system | |
| CN104468304A (en) | Method for synchronizing pool element status information, pool registrator and pool elements | |
| CN102938732B (en) | A kind of method and device improving double-tunnel reliability | |
| CN108600073B (en) | Dynamic tunnel establishment method and device | |
| EP3079327B1 (en) | Information transmission method, device and system | |
| CN106330547B (en) | Method and device for forwarding redundant link data packet |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CP03 | Change of name, title or address | ||
| CP03 | Change of name, title or address |
Address after: 310052 Binjiang District Changhe Road, Zhejiang, China, No. 466, No. Patentee after: Xinhua three Technology Co., Ltd. Address before: 310053 Hangzhou hi tech Industrial Development Zone, Zhejiang province science and Technology Industrial Park, No. 310 and No. six road, HUAWEI, Hangzhou production base Patentee before: Huasan Communication Technology Co., Ltd. |