CN1257464C - Blade Server Management System with Hardware Standby Structure - Google Patents

Abstract

A blade-type server management system with a hardware backup structure is used to manage blade-type servers and blade-type network bridges, and includes at least two management blocks and a baseboard. The management blocks detect heartbeat signals in a predetermined time sequence according to the position of the management blocks inserted on the baseboard to form a main management block for direct management and an auxiliary management block for backup management. Between the main management block and the auxiliary management block, a communication interface is used for data synchronization, and a network interface is used for backup. The system can also use a local management computer connected to another communication interface for backup management of the system. The auxiliary management block of the system can be promoted to the main management block when the main management block fails, so the stability and efficiency of the system can be increased.

Description

Blade Server Management System with Hardware Standby Structure

technical field

The present invention relates to a blade server management system, in particular to a blade server management system with a hardware standby structure.

Background technique

Due to the advancement of technology and the increasing dependence on computer systems. Therefore, the market has higher and higher requirements for the computing power of the server system, but as the available space becomes narrower, the office or factory space occupied by the server system is required to be smaller and smaller. In order to have high stability, the traditional server system is larger than the size of ordinary desktop computers, and in terms of management and space configuration, many problems of server connection and management are also derived. . For general companies, there are as few as two or three servers, but as many as thousands of servers. Therefore, the management and space utilization of servers will become increasingly prominent along with the demand for computers by enterprises.

Blade server is the current trend, which can integrate the hardware of all server systems-including functions such as processors, memory, hard drives, and network connections, into a single expansion card, or so-called "blade" (Blade). In terms of performance, this kind of server has high computing power and high stability, and occupies much less space than traditional servers. In addition, the cost and temperature can be reduced, and various functions and performances are improved compared with traditional servers. The blade server is embedded in the server chassis by inserting the blade, so it is called the blade server. Users can purchase different numbers of servers according to individual needs. It is easy to disassemble and assemble, and it only takes a few seconds to insert a new server. And each server is independent of each other. When the demand increases, you only need to purchase additional servers and embed them. Any insertion or removal will not affect the actions of other servers on the same board. So multiple servers can work in the same server chassis at the same time. It's about the size of a traditional server.

Therefore, blade servers are suitable for enterprise server users such as telecommunications, portal websites, and Internet Service Providers (Internet Services Providers, ISPs), as well as tasks that require a large amount of extremely high-speed computing power, such as meteorological data research, astronomical observation, biotechnology (such as DNA computing), computer animation special effects in the film industry, etc., providing high-speed and stable server devices.

Generally speaking, blade server management can be divided into in-band (In-Band) management and out-of-band (Out-Band) management. The so-called in-band management is structured on the BIOS or operating system (OS) software, and uses software to manage the server. Its disadvantage is that there is no management block. As long as any situation occurs in the server, it will become paralyzed and unable to run. The out-of-band management is to use the management block to manage the server. Although it has a management block, no backup mechanism exists. Since each blade server manages multiple groups of servers at the same time, when the management block fails, the blade server will degenerate from the out-of-band management mode to the in-band management mode. At this time, when any one of the servers is abnormal, the management personnel of the server will no longer be able to obtain the abnormal phenomenon through the management block to take countermeasures and handle it. It causes difficulties for the server administrators in management and affects the stability of the server. Therefore, how to effectively improve the structure of the blade server management system so that the operating status of the server can be controlled by the management personnel and quickly solve the problems on the server is what every information management personnel and enterprise users look forward to.

Contents of the invention

In view of the above technical background, the blade server uses the management block to manage the server. But when each blade server manages several servers to more than 20 servers at the same time, and when the management block fails, the server management personnel can no longer obtain abnormal phenomena through the management block to take countermeasures and processing. It will cause difficulties in server management, thus affecting the stability of server work.

An object of the present invention is to provide a blade server management system with a hardware backup mechanism, which can effectively increase the management stability of the blade server.

Another object of the present invention is to provide a blade server management system, which not only has a hardware backup mechanism, but also utilizes a local management computer for system management.

According to the purpose mentioned above, the present invention is a blade server management system with a hardware backup mechanism, which is used to manage blade servers and blade bridges. The management system includes at least two management blocks, a board, and a memory. Each management block has the ability to generate a Heart Beat signal, and can be used to manage the functions of the above blade servers and blade bridges. Each management block contains two communication interfaces and one network interface. The first communication interface is used for data synchronization among the management blocks, and the second communication interface is connected with a local management computer to directly manage the blade server and the blade network bridge. The network interface is used to connect to the blade bridge for network grouping and information exchange, and also for data synchronization between management blocks when the first communication interface fails.

The base plate has a first management block slot, a second management block slot, and a plurality of blade slots. Wherein, the management block inserted in the first management block slot will automatically form the main management block if there is no heartbeat signal from another management block after waiting for the first predetermined time. The main management block will directly manage the blade server and the blade bridge, and generate a heartbeat signal, which is sent to another management block, and when another management block receives the heartbeat signal, it automatically forms an auxiliary management block. Data synchronization is performed between the management blocks via individual first communication interfaces, so that the auxiliary management block has a backup management mechanism to take over from the main management block. However, if the management block inserted in the second management block slot does not receive any heartbeat signal after waiting for the second predetermined time, the management block will become the master management block. The main management block generates a heartbeat signal and transmits it to another management block, and when the other management block receives the heartbeat signal, it automatically forms an auxiliary management block. Generally speaking, the second predetermined time is set to be longer than the first predetermined time, for example, the first predetermined time is about 3 seconds, and the second predetermined time is about 5 seconds.

The memory is connected to the base board and is used to record the work records of the main management block, and when the auxiliary management block takes over from the main management block, these work records are provided to the new main management block to continue to manage the blade server and the blade bridge .

The above-mentioned network interface can also be directly connected to the network to directly perform network grouping and information exchange. As for the main management block and the auxiliary management block, the main management block and the auxiliary management block can also be changed by the application program. Moreover, the management system also prevents the auxiliary management block from directly managing the blade server and the blade bridge through the isolation of the hardware circuit.

Therefore, the present invention can increase the working stability of the server, so that the server manager can effectively control the operation of the server and obtain the usage status of the server at any time.

Description of drawings

Preferred embodiments of the present invention will be described in more detail in the following description with reference to the following drawings, in which:

FIG. 1 is a schematic diagram of a preferred embodiment of the blade server management system with hardware backup structure of the present invention.

Description of figure number:

102 server 104 bridge

106 The first management block 108 The second management block

110 Network 112 The first local management computer

114 second local management computer 116 memory

118 Substrate 122 Second communication interface

124 network interface 126 first communication interface

132 second communication interface 134 network interface

136 first communication interface

Detailed ways

The spirit of the present invention will be described in detail below with reference to the accompanying drawings and preferred embodiments of the present invention. After understanding the preferred embodiments of the present invention, those who are familiar with the art can change and modify the technology taught by the present invention. without departing from the spirit and scope of the present invention. Referring to FIG. 1 , it is a schematic diagram of a preferred embodiment of a blade server management system with a hardware backup structure according to the present invention. As shown in the figure, the present invention includes a first management block 106 , a second management block 108 , a network bridge 104 , a plurality of servers 102 , a memory 116 for recording system status, and a system substrate 118 . The first management block 106 has a second communication interface 122 , a network interface 124 and a first communication interface 126 , and is connected to the memory 116 through the substrate 118 . Likewise, the second management block 108 has a second communication interface 132 , a network interface 134 and a first communication interface 136 , and is connected to the memory 116 through the substrate 118 .

The blade server management system with hardware backup structure of the present invention uses the first management block 106 and the second management block 108 to manage all the servers 102 and network bridges 104 in the blade server. The actual usage numbers of these servers 102 and network bridges 104 are inserted by users according to work requirements. For example, when a user wants to use two servers and a network bridge, he only needs to insert the two servers and a network bridge into the substrate 118 of the blade server management system with hardware backup structure of the present invention, and the present invention can be used. The management block manages and controls these servers and network bridges. The main function of the server is to provide the function of the computer server in the network, and the network bridge is to provide the function of network signal exchange.

When using the blade server management system with hardware backup structure of the present invention, the first management block 106 and the second management block 108 are respectively inserted into the management block slots of the base plate 118 (not shown). The server 102 and the network bridge 104 to be used are respectively inserted into the blade slots on the substrate 118 . According to the position of the management block slot where the management block is inserted and the heartbeat signal sent by the management block, the first management block 106 and the second management block 108 of the present invention will automatically form a main management block and an auxiliary management block. For example, when the first management block 106 is inserted into the first management block slot, the second management block 108 is inserted into the second management block slot. The first management block 106 inserted in the first management block slot will wait for the first predetermined time, if it does not receive the heartbeat signal sent by other management blocks, then the first management block 106 will automatically form this management block. The main management block of the system. But if within the first predetermined time, the first management block 106 receives heartbeat signals from other management blocks, then the first management block 106 will automatically form the auxiliary management block of the management system, and through the hardware circuit Isolation (Hardware Isolation) design and become monitor mode.

And this moment, the second management block 108 inserted in the second management block slot will wait for the second predetermined time earlier. After exceeding this second predetermined time, if the second management block 108 has not received any If the transmitted heartbeat signal is received, the second management block 108 will automatically form the main management block. However, if the second management block 108 receives heartbeat signals from other management blocks within the second predetermined time, the second management block 108 will automatically form an auxiliary management block of the management system, and through the hardware circuit The design of isolation becomes monitor mode. Generally speaking, in a preferred embodiment of the present invention, the first predetermined time is about 3 seconds, and the second predetermined time is slightly longer than the first predetermined time, about 5 seconds. Therefore, the first management block 106 inserted in the first management block slot will wait for about 3 seconds to determine whether there are other main management blocks in the system that are already carrying out the work of the main management block and sending a heartbeat signal, if not, it will be performed by The first management block 106 forms the main management block. And the second management block 108 that is inserted in the second management block slot will wait for about 5 seconds to determine that there are no other main management blocks in the system after performing management work and send a heartbeat signal, automatically by the second management block. Block 108 forms the main management block. Therefore, the management block of the present invention can automatically form the main management block and the auxiliary management block according to the predetermined time difference.

However, in the blade server management system with a hardware backup structure of the present invention, if no management block is inserted into the first management block slot during actual use, then according to the aforementioned judgment logic, insert the second management block into the second management block slot. The management block in the slot will automatically form the main management block after waiting for about 5 seconds.

When the blade server management system with hardware backup structure of the present invention clearly completes the setting of the main management block and the auxiliary management block, the main management block will be responsible for the management of the system and record all the states of the system to the system substrate In the memory 116, such as Electrically Erasable Programmable Read-Only Memory (Electrically Erasable Prgorammable Read-Only Memory, EEPROM). The auxiliary management block enters the monitoring mode. When the main management block fails, the auxiliary management block will take over the management work and become the main management block automatically. At this time, the new main management block will read the system status record when the original main management block fails from the memory 116, and continue the management work of the whole system. Because the auxiliary management block automatically forms a new main management block when the main management block stops working, the blade server management system of the present invention has a hardware backup structure, so it will not lose control of the system when the original main management block stops working. management work.

After the auxiliary management block forms a new main management block, the failed management block can be directly taken out from the management block slot of the blade server, and can be repaired or replaced with a new management block. When the new management block is inserted into the original first management block slot, the second management block 108 located in the second management block slot has formed the main management block at this time and sends out a heartbeat signal. Therefore, the new first management block 106 located in the slot of the first management block will automatically form an auxiliary management block, which is on standby at any time, and takes over the management work when the main management block fails. Therefore, the blade server management system with hardware backup structure of the present invention can realize blade server management work without management dead angle.

The main management block and the auxiliary management block perform important data synchronization via their respective first communication interfaces 126 and 136 . And through the automatic isolation design of the hardware circuit, the auxiliary management block will not manage the server and the network bridge. The auxiliary management block only uses the first communication interfaces 126 and 136 to obtain the data of the main management block, such as Media-Access Control (MAC) address and Chassis ID of the LAN. The main management block also obtains the data of the auxiliary management block through the first communication interfaces 126 and 136, for example, the MAC address of the Field Replacement Unit (Field Replacement Unit, FRU) and the LAN.

Since the auxiliary management block does not manage the server and the network bridge, it will not cause confusion in management and decrease in efficiency. The data of the main management block is obtained from the first communication interface 126 and 136, so when the main management block fails, or the server administrator requests to change the main management block through the application program, the auxiliary management block can stop the heartbeat signal at the main management block After transmission, it will automatically become the new master management block and send out information to update the relevant data of the system without data loss, and the server administrator also uses this new master management block to manage the system immediately, so there will be no management problems. vacancy.

In order to further ensure that the data and information between the auxiliary management block and the main management block can maintain consistency, so that the auxiliary management block can successfully replace the work of the main management block. Between the management blocks of the hardware backup system managed by the blade server of the present invention, in addition to updating data with the first communication interfaces 126 and 136, network interfaces 124 and 134 are also used to form another standby data exchange circuit . That is to say, when the first communication interfaces 126 and 136 between the management blocks fail due to hardware or other reasons, the network interfaces 124 and 134 will immediately become backup data exchange circuits between the management blocks. The network interfaces 124 and 134 also provide the network connection of the blade server management system with hardware backup structure of the present invention, so as to exchange network packets. The management system of the present invention is connected to the network. On the one hand, the network bridge 104 can be used to reduce the number of cables between the systems. On the other hand, the network 110 can also be directly connected to the network interfaces 124 and 134 on the management block, directly through each Individual network interfaces on each management block for system management and control.

Simultaneously each management block of the blade type server management system with hardware standby structure of the present invention also comprises the second communication interface 122 and 132, for example, for the management communication interface, directly with the first local management computer 112 and the second local management A computer 114 is connected. Therefore, the administrator can also use the local computer to directly manage and control the system of the management block, making the system of the present invention more secure and stable.

Therefore, the blade server management system with hardware backup structure of the present invention utilizes the isolation design of the hardware to automatically isolate the auxiliary management block from the management mechanism, so that the auxiliary management block will not affect the management of the server and the network bridge, and then utilizes the communication The interface and the network interface connect the two management blocks to maintain the synchronization of important data. Therefore, the auxiliary management block can be prepared for backup at any time, and can take over as the main management block at any time, so that the management of the server is more secure and correct. The server management personnel can replace the damaged main server, and then insert it to automatically form a new auxiliary management block. Therefore, the hardware backup structure managed by the blade server of the present invention can replace damaged hardware devices without affecting normal operation at all. Furthermore, due to the use of the spare data exchange capability of the network interface, the data synchronization of the two management blocks is safer and more stable. Therefore, server administrators do not need to worry about the management crisis caused by the failure of a single hardware device, but only need to replace the damaged hardware when the system notifies that the hardware is damaged. Because the backup system has already replaced the original management and communication system immediately, the server management work has truly reached the realm of low-risk management. Moreover, the management block of the present invention can also be directly connected to the network, and connected to the local management computer through the communication interface, so that the stability and safety of the blade server management system with hardware backup structure of the present invention are further improved. The present invention does not limit the use of only two management blocks, and the number of management blocks can be determined according to user requirements.

As understood by those familiar with the art, what is described above is only a preferred embodiment of the present invention, and is not intended to limit the claims of the present invention; all other equivalents that do not depart from the spirit disclosed by the present invention are completed Any change or modification should be included in the appended claims.

Claims (20)

1 one kinds of blade type server management systems with hardware alternate configuration are used for managing blade type server and blade type bridge, and wherein, this blade type server management system with hardware alternate configuration comprises at least:

A substrate has a plurality of insert pocket, the first management piece slot and the second management piece slot, and wherein, these a plurality of insert pocket are used for connecting this blade type server and this blade type bridge;

The first management piece is connected to this first management piece slot, after waiting for for first schedule time, when not receiving first heartbeat signal, forms the main management piece, with control and this blade type server of management and this blade type bridge, and produces second heartbeat signal;

The second management piece is connected to this second management piece slot, after receiving this second heartbeat signal, forms auxiliary management piece, obtains data from this main management piece, to carry out the standby management work of this blade type server and this blade type bridge; And

Storer is connected to this substrate, is used for writing down the job record of this main management piece.

2. the blade type server management system with hardware alternate configuration as claimed in claim 1, wherein, the second above-mentioned management piece is after waiting for for second schedule time, when not receiving this second heartbeat signal, form new main management piece with replacing old main management piece automatically, and from this storer, obtain the job record of old main management piece, to continue control and this blade type server of management and this blade type bridge, and produce this first heartbeat signal, and this storer continues the job record of this new main management piece of record.

3. the blade type server management system with hardware alternate configuration as claimed in claim 2, wherein, the first above-mentioned management piece, when within this first schedule time, receiving this first heartbeat signal, to form new auxiliary management piece automatically, and obtain the data of this new main management piece, to carry out the standby management work of this blade type server and this blade type bridge.

4. the blade type server management system with hardware alternate configuration as claimed in claim 2, wherein, the second above-mentioned schedule time is greater than this first schedule time.

5. the blade type server management system with hardware alternate configuration as claimed in claim 4, wherein, the second above-mentioned schedule time is about 5 seconds, and this first schedule time is about 3 seconds.

6. the blade type server management system with hardware alternate configuration as claimed in claim 1 wherein, between above-mentioned main management piece and this auxiliary management piece, also utilizes communication interface to carry out the work of data sync.

7. the blade type server management system with hardware alternate configuration as claimed in claim 6, wherein, between above-mentioned main management piece and this auxiliary management piece, when this communication interface can't be carried out the work of this data sync, also utilize network interface to carry out the work of this data sync.

8. the blade type server management system with hardware alternate configuration as claimed in claim 7, wherein, the work of above-mentioned data sync comprises: this auxiliary management piece is obtained the medium access control address of Local Area Network of this main management piece and the data of machine plate identifier.

9. the blade type server management system with hardware alternate configuration as claimed in claim 8, wherein, the work of above-mentioned data sync also comprises: this main management piece is obtained the data of the medium access control address of this place substituting unit of assisting the management piece and Local Area Network.

10. the blade type server management system with hardware alternate configuration as claimed in claim 1, wherein, above-mentioned main management piece and this auxiliary management piece, also can require to change this main management piece by application program, make this main management piece become this auxiliary management piece, and should become this main management piece by auxiliary management piece.

11. the blade type server management system with hardware alternate configuration as claimed in claim 1 also via the isolation of hardware circuit, makes this auxiliary management piece can not carry out the direct management work of this blade type server and this blade type bridge.

12. the blade type server management system with hardware alternate configuration as claimed in claim 1, wherein, above-mentioned auxiliary management piece, when this main management piece lost efficacy, the function of this main management piece will be replaced at once, and send the information update system related data, make server administrators proceed the direct management work of this blade type server and this blade type bridge.

13. the blade type server management system with hardware alternate configuration as claimed in claim 1, wherein, above-mentioned main management piece and this auxiliary management piece, has the supervisory communications interface respectively, directly connect the local management computer, carry out the standby management work of this blade type server and this blade type bridge.

14. the blade type server management system with hardware alternate configuration comprises at least:

At least one blade type server, the function of the device of providing services on the Internet;

At least one blade type bridge, the function that provides network signal to exchange;

Two management pieces have the function that produces heartbeat signal and manage these blade type servers and these blade type bridges, and wherein each this management piece also comprises:

First communication interface, carrying out the data sync of these management between pieces,

The second communication interface connects the local management computer, has the function of these blade type servers of direct management and these blade type bridges, with

Network interface is connected to these blade type bridges, carries out network packet and message exchange;

A substrate has the first management piece slot and the second management piece slot, is used for connecting these two management pieces, and a plurality of insert pocket, is used for connecting these blade servers and these blade type bridges, wherein,

When this management piece that is inserted in this first management piece slot, wait for after one first schedule time, do not receive first heartbeat signal that is transmitted by another this management piece that is inserted in this second management piece slot, then be inserted in this management piece of this first management piece slot, form a main management piece, this main management piece is directly managed these blade type servers and these blade type bridges, and produce second heartbeat signal, be sent to another this management piece that is inserted in this second management piece slot, and after another this management piece receives second heartbeat signal, form an auxiliary management piece, should auxiliaryly manage piece and this main management piece,, carry out this data sync via this first communication interface separately, should auxiliaryly manage piece and carry out the standby management work of these blade type servers and this blade type bridge, with

When this management piece of another piece that is inserted in this second management piece slot, wait for after second schedule time, do not receive this second heartbeat signal that is transmitted by this management piece that is inserted in this first management piece slot, then be inserted in another this management piece of this second management piece slot, replace this main management piece, form new main management piece, this new main management piece is managed these blade type servers and these blade type bridges, and produce first heartbeat signal, be sent to this first management piece slot, and this management piece that is inserted in this first management piece slot is when receiving this heartbeat signal, form new auxiliary management piece, should newly assist management piece and this new main management piece, via this first communication interface separately, carry out this data sync, this is assisted management piece and carries out the standby management work of these blade type servers and these blade type bridges, wherein, this second schedule time is greater than this first schedule time; And

Storer, be connected to this substrate, be used for writing down the job record of this main management piece, wherein, when this new main management piece replaces this main management piece, this new main management piece is by obtaining this job record, to continue these blade type servers of management and these blade type bridges, this job record of this storer and this new main management piece of continuation record in this storer.

15. the blade type server management system with hardware alternate configuration as claimed in claim 14, wherein, between above-mentioned main management piece and this auxiliary management piece, when this first communication interface can't be carried out the work of this data sync, also utilize this network interface to carry out the work of this data sync.

16. the blade type server management system with hardware alternate configuration as claimed in claim 15, wherein, the work of above-mentioned data sync comprises this auxiliary management piece and obtains the medium access control address of Local Area Network of this main management piece and the data of machine plate identifier.

17. the blade type server management system with hardware alternate configuration as claimed in claim 16, wherein, the work of above-mentioned data sync comprises the data that this main management piece is obtained the medium access control address of this place substituting unit of assisting the management piece and Local Area Network.

18. the blade type server management system with hardware alternate configuration as claimed in claim 14, wherein, above-mentioned network interface is direct interconnection network also, to carry out this network packet and message exchange.

19. the blade type server management system with hardware alternate configuration as claimed in claim 14, wherein, above-mentioned main management piece and this auxiliary management piece, also can require to change this main management piece by application program, make this main management piece become this auxiliary management piece, and should become this main management piece by auxiliary management piece.

20. the blade type server management system with hardware alternate configuration as claimed in claim 14 also via the isolation of hardware circuit, makes this auxiliary management piece can not carry out the direct management work of these blade type servers and these blade type bridges.

Images