JP2011039920A - System and method for monitoring server - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a server monitoring system which is installed in a server system, wherein a plurality of mother boards sharing equipment such as a power source and a fan module are mounted, and ican monitor/control the server system (in particular, shared modules such as the power source and equipment) without addition of dedicated hardware and firmware. <P>SOLUTION: A monitoring controller 4 and a monitoring controller 5 preliminarily hold MAC address information of each other, and the monitoring controller 4 broadcasts packet data including MAC address information via a network 10 to obtain an IP address corresponding to this MAC address from the monitoring controller having the MAC address. Furthermore, the monitoring controller 4 issues an IPMI over LAC command to the monitoring controller 5 by the obtained MAC address to obtain state information of fans 8 and 9 from the monitoring controller 5. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a server monitoring system and a server monitoring method, and more particularly, to a server monitoring system having a configuration in which a plurality of power supplies and fan modules are connected to one motherboard and these are integrated into one chassis. The present invention relates to a server monitoring method.

In the conventional server system, a plurality of power supplies and fan modules are connected to one motherboard, and these are integrated into one chassis. For this reason, when a plurality of such server systems are installed and managed in a data center or the like, a large installation space is required. In addition, when the power consumption of each of the plurality of server systems is summed, there is a problem that a large amount of power is required as a whole.
For this reason, as a means for solving such a problem, a server system (so-called “blade server” or the like) in which a power supply or a fan module is shared by a plurality of motherboards and incorporated in one chassis is used as an existing system. A server monitoring system corresponding to the server system has been devised.

As a server monitoring system, for example, Patent Document 1 intends to detect a failure of a fan of an information terminal device and notify a failure diagnosis result to a server device. A technology is disclosed in which when a fan failure is detected from the ambient temperature, a transmission / reception unit connected to the server device via a communication network notifies the server device of the occurrence of the failure.
Further, for example, Patent Document 2 discloses a power system as a distributed control monitoring device that enables data for monitoring and control to be transmitted via different transmission paths and that simplifies the configuration and reduces costs. In a configuration in which communication conforming to TCP / IP is performed between a plurality of protection relay control devices to be protected and a monitoring control device via a network, a technique for clustering the monitoring control devices and assigning an IP address is disclosed. ing.

  Further, in Patent Document 3, as a blade server management system, a management subsystem that is provided for each blade and performs remote control of the blade via a local network and the backboard are provided, and at least one is provided. A configuration comprising: a monitoring control subsystem that monitors and controls the operation parameters of the blade server according to a command of the main management subsystem; and a communication subsystem that is provided on the backboard and that connects the local network so as to be able to communicate with an external network A system is disclosed.

JP 2005-4675 A JP 2005-168144 A JP 2008-165803 A

  However, in the existing server system described in the background art, as described above, a power source and a fan module are connected to one motherboard, and these are integrated into one chassis. When multiple server systems are installed and managed in a data center or the like, a large installation space is required, and when the power consumption of each of these server systems is summed, a large amount of power is required as a whole. Therefore, a server monitoring system has been devised that supports a server system (so-called “blade server”, etc.) configured to share a power supply and a fan module with a plurality of motherboards and incorporate them into one chassis. . However, such an existing system still has the following problems.

The first problem is that it is necessary to mount a dedicated motherboard to monitor the shared power supply and fan module, and a dedicated board ("backplane board" for connecting this power supply and fan module etc. Need to be implemented), which complicates the hardware configuration and results in an overall expensive system.
The second problem is that dedicated firmware is required to control the hardware described in the first problem.
The third problem is that in order to remotely monitor and control such a device via a network, a monitoring application (software) that operates on a monitoring terminal such as a PC also depends on the device configuration. It is necessary to make it a dedicated program.

As described above, the technique disclosed in Patent Document 1 described above is intended to detect a failure of a fan of an information terminal device and notify a server apparatus of a failure diagnosis result. As described above, the status of other server devices is not grasped between the server devices, and the overall status of the server system is not grasped as a whole. Therefore, the status of the entire server system can be easily grasped from the external terminal device as well. It is not possible.
In addition, as described above, the technique disclosed in Patent Document 2 described above enables data for monitoring and control to be transmitted via different transmission paths and simplifies the configuration to reduce the cost. However, the use of the communication network is limited between the protection relay control device and the monitoring control device, and in order to grasp the status of other server devices between server devices as in the present invention. not being used.

  Further, as described above, the technique disclosed in Patent Document 3 includes a management subsystem that is provided for each blade and performs remote control of the blade via a local network. One is that the main management subsystem provided controls the operation of the supervisory control subsystem, but as in the present invention, a plurality of server devices having completely the same configuration and completely the same function. It does not grasp the status of other server devices between them, and requires at least one main management subsystem. Therefore, the structure becomes complicated and the cost is increased. Requires system-specific software.

The present invention has been made in view of the above-described conventional problems, and is installed in a server system in which a plurality of motherboards that share devices such as a power supply and a fan module are mounted, and has dedicated hardware and firmware. It is an object of the present invention to provide a server monitoring system that can realize monitoring and control of a server system (particularly, a shared module such as a power supply and devices) without adding a password.
Another object of the present invention is to monitor the entire system by an application having a function capable of integrally monitoring various types of server systems through a uniform interface. To provide a system.

  In order to solve the above-mentioned problems, a server monitoring system according to the present invention is a server monitoring system for managing a server system in which a plurality of motherboards on which a shared module is mounted is installed, and each of the plurality of motherboards has a monitoring controller. And a distributed control means for distributing and monitoring the shared module to each of the plurality of monitoring controllers, and installed in each of the plurality of monitoring controllers. Communication means for communicating with each other via an external network, and each of the plurality of monitoring controllers is in charge of and held by another monitoring controller via the communication means. As well as grasping the status information of the shared module, the service is based on the grasped status information. Characterized in that to grasp the state of the entire system.

  The server monitoring method according to the present invention is a server monitoring method for managing a server system in which a plurality of motherboards on which a shared module is mounted is mounted, wherein a monitoring controller is mounted on each of the plurality of motherboards, and A distributed control step for distributing and monitoring the shared module to each of the plurality of monitoring controllers, and each of the plurality of monitoring controllers is installed in each of the plurality of monitoring controllers. A communication step for communicating with each other via the communication step, and each of the plurality of monitoring controllers is in charge of and held by the other monitoring controller via the communication step. And the server system as a whole based on the grasped status information. Characterized in that to grasp the state.

  As described above, according to the server monitoring system of the present invention, in a server monitoring system that manages a server system in which a plurality of motherboards with shared modules are mounted, without adding dedicated hardware or firmware, There is an effect that it is possible to provide a server monitoring system capable of performing monitoring / control work of a server system (particularly, a shared module such as a power supply and devices).

It is a lineblock diagram showing the whole server monitoring system composition concerning an embodiment of the present invention. It is explanatory drawing which shows the tables which the monitoring controller 4 accesses for monitoring. It is explanatory drawing which shows the tables which the monitoring controller 5 accesses for monitoring. It is a flowchart figure which shows operation | movement of the server monitoring system which concerns on embodiment of this invention.

  The server monitoring system of the present invention is a system in which a plurality of motherboards are incorporated, and as a means for monitoring and controlling the shared power supply and fan module, the monitoring and control of the power supply and fan module are shared by each of the plurality of motherboards. It is characterized by making it. Each of the monitoring / control controllers (BMC) installed on each motherboard can monitor the entire system as a whole by communicating with each other via a network. For this reason, it is possible to eliminate the need for dedicated hardware for integrated monitoring required by an existing server monitoring system.

That is, in the server monitoring system of the present invention, the server monitoring controller extracts the MAC address (Media Access Control address) information of the monitoring controller of the same group with reference to the group table, and includes the MAC address information on the network. Send data over the network.
If the monitoring controller that has received the packet is a monitoring controller in the same group, the IP address information corresponding to the MAC address is returned to the monitoring controller that sent the MAC address information packet data.

  Furthermore, while issuing the "IPMI over LAN" command (command conforming to the standard interface specification) to the monitoring controller of the IP address destination acquired by the monitoring controller, the status of the monitoring target module under this controller is acquired, Add (reflect) the acquired monitoring module status information to the monitoring target table. For all accesses to the monitoring controller from the outside, all the monitoring module status information reflected in the monitoring target table is returned.

Hereinafter, an embodiment of a server monitoring system of the present invention will be described in detail with reference to the drawings.
FIG. 1 is a configuration diagram showing the overall configuration of a server monitoring system according to an embodiment of the present invention.
In the figure, the server monitoring system 1 of this embodiment includes a motherboard 2 and a motherboard 3 as two motherboards, a monitoring controller 4 and a monitoring controller 5 mounted on the motherboards, and a power source shared by the two motherboards. 6, and a fan 7, a fan 8, and a fan 9 that are shared in the server system 1. The monitoring controller 4 and the monitoring controller 5 are connected to be able to communicate with each other via an intranet network 10.

In general, the motherboard 2 and the motherboard 3 of the server monitoring system 1 include other components (not shown) (for example, components that realize functions as an original server system).
In this embodiment, the monitoring controller 4 and the monitoring controller 5 are provided. However, in general, in the server monitoring system according to the present invention, the number of installed monitoring controllers is an arbitrary number of 2 or more.

In this embodiment, the type of device controlled by the monitoring controller 4 and the monitoring controller 5 is a fan. However, in general, in the present invention, the type of device controlled by the monitoring controller is arbitrary, and the number of devices installed is also the same. Is optional.
Furthermore, in this embodiment, one power source is shared by the mother board 2 and the mother board 3. However, in general, in the present invention, the number of power supplies and the arrangement configuration thereof are arbitrary.
The motherboard 2 supplies power for driving the fan 7. The mother board 3 supplies power for driving the fan 8 and the fan 9.

Hereinafter, functions of each component included in the server monitoring system of the present embodiment will be described.
The monitoring controller 4 has a function of accessing the fan 7 and the power supply 6 that are directly connected to the controller, and acquiring the state information (normal or abnormal). Similarly, the monitoring controller 5 has a function of accessing the fan 8 and the fan 9 and acquiring state information thereof. Each of the monitoring controller 4 and the monitoring controller 5 has its own MAC address and a corresponding IP address, and is connected to each other via an intranet network 10 so as to communicate with each other.

  The monitoring controller 4 and the monitoring controller 5 hold each other's MAC address information in advance, and the monitoring controller 4 broadcasts (broadcasts) the packet data including the MAC address information via the network 10. An IP address corresponding to the MAC address is acquired from the monitoring controller 5 having the MAC address. Furthermore, the monitoring controller 4 acquires the status information of the fans 8 and 9 from the monitoring controller 5 by issuing the above-mentioned “IPMI over LAN” command to the monitoring controller 5 with the acquired IP address.

As a result, the monitoring controller 4 holds the status information of the fans 8 and 9 under the monitoring controller 5 in addition to the fans 7 and the power supply 6 under the monitoring controller 4. The status information of the entire server monitoring system 1 including the configured server system is held.
The state of the entire server monitoring system 1 including the motherboard 2 and the server system composed of the motherboard 3 by accessing the monitoring controller 4 via the network 10 from an external monitoring terminal device or the like (not shown). Information can be confirmed.

FIG. 2 is an explanatory diagram showing tables that the monitoring controller 4 accesses for monitoring.
FIG. 3 is an explanatory diagram showing tables that the monitoring controller 5 accesses for monitoring.
2 and 3, reference numeral 41 denotes a self-network setting table, reference numeral 42 denotes a group table, and reference numeral 43 denotes a monitoring target table.
The monitoring controller 5 also corresponds to the tables shown in FIGS. 2 and 3 and is accessed by the monitoring controller 5.
In the self-network setting tables 41 and 51 and the group tables 42 and 52, “MAC address” and “IP address” are provided as table items. The monitoring target tables 43 and 53 are provided with “monitoring controller”, “monitoring target”, and “state” as table items.
The role of each table will be described in conjunction with the description of the operation described later.

(Description of operation)
FIG. 4 is a flowchart showing the operation of the server monitoring system according to the embodiment of the present invention.
Hereinafter, the operation of the server monitoring system according to this embodiment will be described with reference to FIGS.
(Step A1)
First, in step A1, the monitoring controller 4 (see FIG. 1) mounted on the motherboard 2 accesses the fan 7 and the power source 6 (see FIG. 1) connected to the controller, and state information (normal or normal). Get abnormal). Similarly, the monitoring controller 5 (see FIG. 1) acquires state information of the fan 8 and the fan 9 (see FIG. 1).
(Step A2)
Next, in step A2, the monitoring controller 4 adds the acquired state information to the monitoring target table 43 (see FIG. 2). At this time, the monitoring target table 43 does not reflect the state information of the fans 8 and 9 as shown in FIG. Similarly, the monitoring controller 5 adds the acquired status information of the fans 8 and 9 to the monitoring target table 53 (see FIG. 3).

(Step A3)
Next, in step A3, the monitoring controller 4 extracts the MAC address information (MAC # 2 in this example) from the group table 42 holding the MAC address information of the monitoring controllers belonging to the same server monitoring system 1. (See FIG. 2). The MAC address information in the group table 42 is information set when the server monitoring system 1 is manufactured. At the time of manufacturing, the group table 42 is not set with an IP address corresponding to the MAC address information (MAC # 2).
(Step A4)
Next, in step A4, the monitoring controller 4 uses the MAC address information (see FIG. 2) of the own network setting table 41 (see FIG. 2) that holds the extracted MAC address information (MAC # 2) and the network information of the monitoring controller 4 itself. Packet data including MAC # 1) is generated and broadcast (simultaneously transmitted) on the network 10 (see FIG. 1). The broadcast packet data reaches the monitoring controller 5 via the network 10.

(Step A5)
Next, in step A5, the monitoring controller 5 receives packet data.
(Step A6)
Next, in step A6, the monitoring controller 5 confirms that the MAC address (MAC # 2) included in the received packet matches its own MAC address.
(Step A7)
Next, in step A7, the monitoring controller 5 extracts the IP address (IP # 2) corresponding to the MAC address (MAC # 2) from the self network setting table 51 (see FIG. 3), and this IP address (IP The packet data including # 2) is returned to the monitoring controller 4.
(Step A8)
Next, in step A8, the monitoring controller 4 receives the packet data, extracts the IP address information (IP # 2) included in the packet, and stores the IP address information (IP # 2) in the group table 42. IP # 2) is added as an IP address corresponding to MAC # 2.

(Step A9)
Next, in step A9, the monitoring controller 4 issues the above-mentioned “IPMI over LAN” command to the monitoring controller 5 using the IP address information (IP # 2) added to the group table 42. To re-access (step A9, FIG. 1). Here, IPMI is a standard interface specification for managing and monitoring servers without depending on hardware or software. IPMI over LAN accesses this interface via a network. It is to be.
(Step A10)
Next, in step A10, the monitoring controller 5 extracts the status information (normal or abnormal) of the fans 8 and 9 from the monitoring target table 53 (see FIG. 3), and sends it back to the monitoring controller 4.

(Step A11)
Finally, in step A11, the monitoring controller 4 adds the status information of the fans 8 and 9 to the monitoring target table 43 (see FIG. 2).
The operations of the monitoring controller 4 and the monitoring controller 5 in the above steps A1 to A11 can be interchanged.
Through a series of operations over the above steps A1 to A11, the monitoring controller 4 and the monitoring controller 5 can grasp the state of the devices in the server monitoring system 1 including the partner's jurisdiction range.

  Further, the status information of the entire server monitoring system 1 can be confirmed by accessing the monitoring controller 4 from the monitoring terminal device outside the monitoring controller 4 via the network 10. That is, by this operation, the status information of the entire server monitoring system 1 can be confirmed by accessing either the monitoring controller 4 or the monitoring controller 5 via the network 10 from an external monitoring terminal or the like. In this case, it is not necessary to access both the monitoring controller 4 and the monitoring controller 5.

  As described above, the server monitoring system 1 according to the present embodiment is like an existing server monitoring system in a hardware configuration in which a plurality of independent server systems are mounted in the same apparatus and a power source and a fan module are shared. In addition, it is no longer necessary to install dedicated hardware (backplane or dedicated communication bus for connecting modules), dedicated firmware, etc. to integrate and monitor these modules. Since the system can be configured only by mounting a basic network communication function for another controller in the control controller, it is possible to provide a very inexpensive system.

  Also, when using integrated monitoring software that monitors each of these server systems from outside, the existing server monitoring system is dedicated to show information from each of multiple monitoring / control controllers as information in the same system. However, in the server monitoring system 1 according to the present embodiment, the function of the monitoring / control controller described above makes it uniform even from the outside. Because the procedure can monitor the status of the server system, there is no need to install dedicated software on the management software side, such as an existing server monitoring system, and even when monitoring different types of server systems, the user Standardized without changing the interface Possible and it is effective in the management and administration work is facilitated.

(Other embodiments)
In the above-described embodiment, a configuration in which two motherboards are mounted in the server system has been described. However, in general, the present invention is applied to a server system including a plurality of motherboards in which three or more motherboards are mounted. Can also be applied. In this case, in the group table 42 (FIG. 2), MAC address information may be set in advance in an entry having the number of entries of “motherboard mounting number−1”.

  A program for executing at least a part of the processing of each component of the server monitoring system according to the present invention by computer control and causing the computer to execute the above processing according to the procedure shown in the flowchart of FIG. It may be stored and distributed in a computer-readable recording medium such as a semiconductor memory, a CD-ROM or a magnetic tape. A computer including at least a microcomputer, a personal computer, and a general-purpose computer may read and execute the program from the recording medium.

  The present invention can be applied to the construction of a server monitoring system, and in particular, a server system in a form in which a plurality of power supplies and fan modules are connected to one motherboard and these are integrated into one chassis. It is suitable for providing a server monitoring system for monitoring.

1 Server monitoring system 2, 3 Motherboard 4, 5 Monitoring controller
6 Power supply 7, 8, 9 Fan 10 Intranet network

Claims (8)

A server monitoring system for managing a server system comprising a plurality of motherboards equipped with shared modules,
A monitoring controller mounted on each of the plurality of motherboards, and distributed control means for performing a task of monitoring and controlling the shared module in a distributed manner to each of the plurality of monitoring controllers;
A communication means that is installed in each of the plurality of monitoring controllers, and each of the plurality of monitoring controllers communicates with each other via an external network;
Each of the plurality of monitoring controllers grasps the status information of the shared module that is in charge of and held by another monitoring controller via the communication means, and based on the grasped status information, the server A server monitoring system characterized by grasping the state of the entire system.   The communication means installed in each of the plurality of monitoring controllers communicates with each of the other monitoring controllers via the external network in order to communicate with each other's MAC address (Media Access Control). 2. The server monitoring system according to claim 1, wherein address information is held.   Each of the monitoring controllers acquires corresponding IP address information from the MAC address information of each of the other monitoring controllers, and uses the IP address information to designate a command (standard interface specification) called “IPMI over LAN”. The status information related to the shared module held by each of the other monitoring controllers is acquired from each of the other monitoring controllers as a response. 2. The server monitoring system according to 2.   Each of the monitoring controllers grasps and holds the status information of the entire server system by merging the status information acquired from each of the other monitoring controllers with the status information held by itself. The server monitoring system according to claim 3.   5. The server monitoring system according to claim 4, wherein each of the monitoring controllers responds with status information of the entire server system when accessed from an external monitoring terminal device.   6. The server monitoring system according to claim 5, wherein each of the monitoring controllers can provide status information of the entire server system to the external monitoring terminal device through a uniform interface.   The server monitoring system according to claim 1, wherein the shared module includes a power supply module and a fan module. A server monitoring method for managing a server system comprising a plurality of motherboards equipped with shared modules,
A distributed control step in which a monitoring controller is mounted on each of the plurality of motherboards, and a task of monitoring and controlling the shared module is distributed to each of the plurality of monitoring controllers.
A communication step installed in each of the plurality of monitoring controllers, each of the plurality of monitoring controllers communicating with each other via an external network,
Each of the plurality of monitoring controllers grasps the status information of the shared module which is in charge of and held by another monitoring controller through the communication step, and the server based on the grasped status information. A server monitoring method characterized by grasping the state of the entire system.

Images