JP2005234749A - Backup system and backup method for server - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To smooth server allocation, in a shared backup center backing up a plurality of user systems for reduction of holding cost of a backup system for in-disaster exclusive use. <P>SOLUTION: This backup system has: a server resource management database storing a management table including priority for allocating a shared server of a shared server resource group having a backing-up database to back up the plurality of user systems to the user systems; and a server allocating system for allocating the shared server to the user systems on the basis of the management table. A plurality of shared servers are used in common use by the plurality of user systems even out of time of the backup, and the server allocating system selects the user system on the basis of the priority of the management table, and allocates it to the shared server, at the time of the backup. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a server backup system and a backup method.

Conventionally, a server system that needs to be backed up in the event of a disaster reserves a dedicated server model for disaster backup in advance as a disaster backup. The system is backed up using the reserved backup dedicated server.
JP 2001-34595 A

  In the disaster backup system as described above, a backup dedicated server secured in advance is used to perform system backup at the time of disaster. Therefore, since a dedicated server that is used only at the time of a disaster is introduced, there has been a problem that the dedicated server becomes idle assets and the introduction cost increases.

  In order to solve the above problems, a server backup system according to the present invention stores a management table including a priority for assigning a shared server of a shared server resource group having a backup database to the user system in order to back up a plurality of user systems. A backup server system, and a server allocation system for allocating the shared server to the user system based on the management table, wherein the plurality of shared servers include the backup server The server allocation system selects the user system and allocates the shared server based on the priority of the management table at the time of backup. To.

The management table may have the priority assigned to each user who uses a plurality of the user systems.
Furthermore, the management table may be assigned the priority in units of groups including a plurality of the user systems.
Furthermore, the shared server resource group may be further provided.

  In order to solve the above problems, the server backup method according to the present invention is such that a shared server of a shared server resource group having a backup database for backing up a plurality of user systems is shared by the plurality of user systems other than at the time of backup. The user system is selected and the shared server is allocated based on the priority of the management table including the priority to allocate the shared server to the user system at the time of the backup.

  The shared server resource group is originally used for backup of a plurality of user systems in preparation for a disaster. That is, the shared server resource group is shared by a plurality of user systems instead of the conventional redundant configuration in which equivalent server systems are individually prepared for each user system. As a result, the number of backup servers can be reduced. As a result, the scale of the entire backup system can be reduced, and the introduction cost can be reduced.

  In addition, the shared server resource group can be used for each user system even during normal times other than during backup due to a disaster or the like, and can be used effectively. As a result, the entire backup system can be prevented from being idled and effectively utilized.

  Further, the server allocation system allocates the shared server to the user system based on the management table of the server resource management database. Therefore, even if shared use is concentrated and contention occurs between user systems, smooth shared use without contention can be achieved. In particular, a reliable backup system can be secured in the event of a system failure due to a disaster or the like.

=== Aspect of Shared Backup Center ===
The use mode of the shared backup center by a plurality of user systems is shown in the schematic diagram of FIG. Each customer (user) has a user system 100. Each user system 100 is connected to each system 210, 240 in the shared backup center through a communication line such as a dedicated line or the Internet. Each of these user systems 100 includes a monitoring system 130 in addition to a plurality of user servers 110 and a database 120, for example. This monitoring system 130 monitors the operation state such as whether each user server 110 has fallen down due to a disaster or the like. The monitoring system 130 is realized by a computer system.

  The shared backup center includes an integrated monitoring system (server allocation system) 210, a server resource management database 220, a server resource management system 230, and a shared server resource group 240. The integrated monitoring system 210 and the server resource management system 230 are realized by a computer system.

  The shared server resource group 240 includes an aggregate of a plurality of shared servers 240a and a backup database 240b for backup. The backup database 240b is configured by a disk array device or the like, and data recorded in the database in each user system 100 is stored as a backup copy. Further, the shared server resource group 240 includes surplus resources including a CPU and a memory in preparation for a shortage of the shared server 240a at the time of backup, as will be described later.

  This shared server resource group 240 is originally used for backup of a plurality of user systems 100 in preparation for a disaster. In other words, the shared server resource group 240 is shared by a plurality of user systems 100 instead of the conventional redundant configuration in which equivalent server systems are individually prepared for each user system 100. As a result, the number of backup servers can be reduced. As a result, the scale of the entire backup system can be reduced, and the introduction cost can be reduced.

  In addition, the shared server resource group 240 can be used for each user system 100 even during normal times other than during a disaster so that it can be used effectively. As a result, the entire backup system can be prevented from being idled and effectively utilized.

  By the way, as described above, the shared server resource group 240 is shared by a plurality of user systems 100. If the shared use is concentrated and the user systems 100 compete with each other, not only the smooth shared use is achieved, but there is a case where a reliable backup system cannot be secured especially when a system down occurs due to a disaster or the like.

  Therefore, in order to achieve smooth shared use without contention, the integrated monitoring system 210 executes system allocation processing based on various management tables pre-registered by contract in the server resource management database 220, and the processing results are sent to the server. Communicate to the resource management system 230. The server resource management system 230 allocates the shared server resource group 240 to each customer who uses each user system 100 according to the transmitted system allocation processing result. The allocated shared server resource group 240 is provided to the customer as a customer system.

=== Allocation of shared server resources ===
A specific shared server resource group 240 allocation process will be described with reference to FIG. First, processing distribution according to the number of customers whose servers are down due to a disaster or the like will be described. As shown in the flowchart of FIG. 2, the integrated monitoring system 210 determines whether or not there are a plurality of customers whose servers are down or a single customer (S210). If it is independent (S210: NO), the process proceeds to execution of an allocation process (described later) shown in the flowchart of FIG. 6 (S220). On the other hand, in the case of a plurality (S210: YES), the integrated monitoring system 210 determines whether or not it is possible to back up with the server resources originally secured when backing up the user systems 100 of a plurality of customers. Specifically, the magnitude relation of the number of user servers 110 that have gone down is confirmed with respect to the number of shared servers 240a pooled for the customer (S230). As a result of the confirmation, if the number of shared servers 240a is equal to or greater than the number of user servers 110 that have gone down, the integrated monitoring system 210 determines that backup is possible with the reserved shared servers 240a, and the flowchart of FIG. The process proceeds to execution of the allocation process shown (S240).

  On the other hand, when the number of shared servers 240a is less than the number of user servers 110 that have gone down, a complete backup is impossible because there are not enough reserved shared servers 240a. That is, a race condition occurs between the user servers 110 that are down. Therefore, the integrated monitoring system 210 moves to the execution of the allocation process of the shared server 240a in order to select the shared server 240a to be provided to the customer and eliminate the race condition (S250).

  The allocation process of the shared server 240a will be described. As shown in the flowchart of FIG. 3, the integrated monitoring system 210 refers to the customer system information management table (management table including priority) stored in the server resource management database 220 (S310).

  This customer system information management table is configured as shown in FIG. 7, for example. As shown in FIG. 7, the contents of the customer system information management table include five levels of customer priority keys (the higher the level, the higher the priority) for each customer (Company A to Company Z), and each customer. User system (A system ~) 100, group numbers to which these user systems 100 belong, five levels of system priority keys for each group number (the higher the level, the higher the priority), and necessary resource information It is attached. This required resource information is a resource required at the time of backup of each user system 100, and its items are the number of CPUs mounted, the number of memories mounted (unit: GB), and the number of disk paths.

  That is, based on this customer system information management table, the integrated monitoring system 210 confirms the customer priority key in the customer system information management table of FIG. 7 in order to prioritize the customers of the user server 110 in a competitive state. To do. Then, it is checked whether or not the customers in the competitive state have the same customer priority key level (S320).

  If the customer priority key stage is the same (S320: YES), the priority order of the customers is not superior or inferior, and the integrated monitoring system 210 moves to the allocation rate calculation process as shown in the flowchart of FIG. 4 (S410). In this calculation process, 100% is divided by the number of customers to be allocated in order to make an even split, and this divided value is used as the allocation ratio. Next, the integrated monitoring system calculates the number of servers that can be allocated for each customer by multiplying the calculated allocation rate by the server resources that are secured as shared (S420). Specifically, for example, in the relationship between two companies, Company A and Company B, where the customer priority key stage is the same “5”, the CPU, memory, And the number of disk paths or I / O paths is multiplied by an allocation rate of 50% (100% / 2 companies) for each of Company A and Company B. The number of shared servers 240a (A to J) that can satisfy the number of CPUs, memories, and disk paths obtained as a result of the multiplication is calculated with reference to the server resource management table of FIG. That is, the shared servers 240aA-J in which the CPU, the memory, and the I / O path in the server resource management table of FIG. 8 satisfy the contents of the multiplication result described above are assigned to the A company and the B company respectively. Select.

  Then, the integrated monitoring system 210 refers to the customer system information management table of FIG. 7 and confirms the system priority key of the customer who is in a competitive relationship. As a result of the confirmation, one of the shared servers 240a (A to J) of FIG. 8 selected in S420 is preferentially assigned to the system group having a high system priority key, and the user system 100 is restored as the customer system. (S430 → S440). Details of this allocation processing will be described later with reference to the flowchart of FIG. This allocation process is executed until the allocation process of the shared server 240a (A to J) selected in S420 reaches the allocation rate of each customer obtained in S410 and cannot be allocated.

  Next, the integrated monitoring system 210 checks whether or not the shared servers 240a (A to J) that can be allocated in S420 remain in an unallocated state (S450). If it does not remain, the assignment process is terminated (S450 → END), and if it remains, the user system 100 that is not assigned in another customer's system is assigned to the system priority key in the descending order as in S430. An allocation process (described later with reference to FIG. 6) is executed for the user systems 100 in the system group (S470 → S480). At the time of this allocation process, when the allocation process is executed for a plurality of user systems 100 within the same system group, the allocation process of the user systems 100 is sequentially performed in the order registered in the customer system management table.

  On the other hand, when the user priorities are different in the determination processing of S320 described above in FIG. 3, as shown in the flowchart in FIG. 5, the integrated monitoring system 210 refers to the customer priority key in the customer system information management table in FIG. Then, a shared resource allocation process (described later with reference to FIG. 6) is executed for a customer with high priority (S510 → S520). Then, the integrated monitoring system 210 checks whether or not the unallocated shared server 240a remains in the shared server resource group after the allocation process (S530). As a result of this confirmation, if there is no unallocated shared server 240a remaining, the process is terminated (S530 → END). If left, the shared server 240a (A to J) is sent to the customer with the next highest customer priority key. Perform the allocation process. Specifically, the allocation process is performed on the user system 100 in descending order of the priority system key (S540 → S550). At this time, if there are a plurality of customers having the same customer priority key stage, the process proceeds to S410 in FIG. 4 (S540 → S410).

  Here, with reference to FIG. 6, the allocation process of the shared server to the user system 100 will be described. The integrated monitoring system 210 extracts the shared server 240a that can satisfy the CPU, memory, and disk path required as the backup system of the user system 100 based on the necessary resource information of FIG. Refer to As a result, the integrated monitoring system 210 is one of the shared servers A to J in which the scale of the CPU, memory, and I / O path in the server resource management table of FIG. To extract. Any one of the extracted shared servers 240a (A to J) is provided as a backup customer system of the user system 100, and the user system 100 is restored (S610 → S620 → S630).

  On the other hand, if any of the shared servers A to J that can be satisfied as the backup system of the user system 100 does not exist, the user system 100 is in the order of registration in the customer system information management table of FIG. With reference to the surplus resource management table of FIG. 9, a process of assigning surplus resources is executed to restore the user system (S610 → S620 → S640 → S650).

  The above-described embodiments of the present invention are intended to facilitate understanding of the present invention, and are not intended to limit the present invention. The present invention can be changed and improved without departing from the gist thereof, and equivalents thereof are also included in the present invention.

It is a schematic diagram which shows the utilization aspect of the shared backup center which concerns on one embodiment of this invention. It is a flowchart which shows the distribution operation | movement of the process according to the number of the customers who the server went down by the disaster etc. by the server backup system which concerns on one embodiment of this invention. It is a flowchart which shows the allocation process of the shared server with respect to the customer who the server went down by the disaster etc. by the server backup system which concerns on one embodiment of this invention. It is a flowchart which shows the calculation process of the allocation rate of a resource in case the stage of a customer priority key is the same by the backup system of the server which concerns on one embodiment of this invention. It is a flowchart which shows the calculation process of the allocation rate of a resource in case the stage of a customer priority key differs by the backup system of the server which concerns on one embodiment of this invention. It is a flowchart which shows the specific shared server allocation process by the server backup system which concerns on one embodiment of this invention. It is a graph which shows the customer system information management table by the backup system of the server which concerns on one embodiment of this invention. It is a chart which shows the server resource management table by the backup system of the server which concerns on one embodiment of this invention. It is a chart which shows the surplus resource management table by the backup system of the server which concerns on one embodiment of this invention.

Explanation of symbols

100 User system 110 User system 120 Database 130 Monitoring system 200 Shared backup center 210 Integrated monitoring system (server allocation system)
220 Server Resource Management Database 230 Server Resource Management System 240 Shared Server Resource Group 240a Shared Server Aggregation 240b Backup Database

Claims (5)

A server resource management database in which a management table including a priority for assigning a shared server of a shared server resource group having a backup database to back up a plurality of user systems to the user system;
A server allocation system for allocating the shared server to the user system based on the management table;
A backup system comprising:
The plurality of shared servers are used jointly by the plurality of user systems other than during the backup,
At the time of the backup, the server allocation system selects the user system and allocates the shared server based on the priority of the management table.
A server backup system characterized by that.   2. The server backup system according to claim 1, wherein the management table is assigned the priority in units of users who use a plurality of the user systems.   2. The server backup system according to claim 1, wherein the priority is assigned to the management table in a group unit including a plurality of user systems.   The server backup system according to claim 1, further comprising the shared server resource group. A shared server of a shared server resource group having a backup database to back up a plurality of user systems is used jointly by the plurality of user systems other than at the time of backup,
Assigning the shared server by selecting the user system based on the priority of the management table including the priority of assigning the shared server to the user system at the time of the backup;
A server backup method characterized by the above.

Images