KR100235884B1 - Method for connecting hard disk and duplexing host in the communication system - Google Patents

Abstract

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a duplexed host of a communication system. In particular, the present invention relates to a hard disk drive quickly without interruption of service when switched from an active host to a standby by host. (Hard Disk Drive) The present invention relates to a hard disk access method of a redundant host in a communication system capable of reconnecting a hard disk drive.

In the case of accessing and using a hard disk as a DAD in a redundant host according to the present invention, by periodically recognizing its own state and relative state and remounting the hard disk of the host which has been switched from the standby state to the operating state, the hard disk at the time of host switching Reconnect the disk and prevent access to the hard disk at the same time by using NVRAM lock and DPRAM access code.Connect the hard disk by resetting the host SCSI bus according to the SCSI identification number classification and host status of the host. The hard disk is accessed from the ROM according to the NVRAM lock code and the DPRAM access code, and the hard disk is accessed from RAM by writing the unlock code to the VNRAM according to the host status. Normal service without interruption of service in the communication system through It can proceed to improve the reliability and efficiency of the communication system.

Description

Hard Disk Connection Method of Redundant Host in Communication System

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a duplexed host of a communication system. In particular, the present invention relates to a hard disk drive quickly without interruption of service when switched from an active host to a standby by host. (Hard Disk Drive) The present invention relates to a hard disk access method of a redundant host in a communication system capable of reconnecting a hard disk drive.

In general, when a host board, which is a main processor unit (MPU), is connected to a hard disk drive that is a direct access device (DAD) in a communication system, the host is a running host and a waiting host to prevent service interruption. Redundant, that is, when a live host fails, the host can be stopped and the host can be started to check and repair the failed host. Even if a host fails, it can be transferred to another host to provide continuous service.

However, when a service failure occurs in the operating host, it takes some time to stop the operation and switch to the waiting host. In order for the waiting host to perform its normal operation, the service is interrupted for some time. There was a disadvantage of this occurring.

As described above, in order to solve the disadvantage that occurs when the redundant host is switched, the present invention, when a hard disk drive is connected to the redundant host in the communication system, an error occurs in the operating host and is transferred to the waiting host. It aims to improve the reliability and efficiency of the communication system by quickly reconnecting the hard disk drive to perform normal service without interrupting service.

1 is a block diagram showing a hardware configuration of a hard disk access device of a redundant host in a communication system according to an embodiment of the present invention.

2 is a block diagram showing a software configuration of a hard disk access device of a redundant host in a communication system according to an embodiment of the present invention.

3 is a flowchart illustrating a hard disk access method of a redundant host in a communication system according to an embodiment of the present invention.

FIG. 4 is a flowchart illustrating a hard disk access operation in accordance with host duplication in the ROM processing unit in FIG. 3. FIG.

FIG. 5 is a flowchart showing a hard disk connection operation according to host duplication at the hard disk connection unit in FIG. 3; FIG.

FIG. 6 is a flowchart illustrating a hard disk connection operation according to host duplication in the RAM processing unit in FIG. 3. FIG.

FIG. 7 is a flowchart showing hard disk reconnection and simultaneous connection prevention operation during host switching by the timer processor in FIG. 3; FIG.

Explanation of symbols on the main parts of the drawings

10, 20: host 11, 21: nonvolatile RAM (NVRAM)

12, 22: DPRAM (Dual Port Memory)

13, 23: SCSI (Small Computer System Interface Unit)

14: CPU

30: RDC (Redundaunt Array of Independent Disks) Disks Controller

40: hard disk connection unit 50: ROM processing unit

60: RAM processing unit 70: Timer processing unit

In order to achieve the above object, a method of accessing a hard disk of a redundant host in a communication system according to an exemplary embodiment of the present invention provides a connection operation between a host and a hard disk in operation by performing a RAM process after performing an ROM process. At the same time, the re-connecting and simultaneous access prevention operations of the standby host and the hard disk are performed.

The ROM processing operation may include: a first process of checking whether a state of the ROM is a host in operation; A second step of recording a lock code and performing a connection prevention operation between the waiting host and a hard disk if its status is the active host; A third step of checking whether a DPRAM code is access enabled if its state is not the operating host; And performing a connection preventing operation between the standby host and the hard disk or performing a connection operation between the operating host and the hard disk when the DPRAM code is accessed. The RAM processing operation may include: a fifth process of checking whether a state of the host is the operating host after the ROM processing operation is performed; And a sixth step of recording an unlock code and performing a reconnection of the standby host and the hard disk when its state is the operating host. The connection operation of the operating host and the hard disk is performed. A seventh step of resetting the SCSI bus only when its state is in operation by classifying the SCSI identification numbers according to the mounted master and slave positions of the host to a target; And an eighth process of mounting the disk driver, the disk buffer cache, and the volumes in order, respectively. The operation of preventing reconnection and simultaneous access between the waiting host and the hard disk is performed by detecting a state transition of the host and waiting. A ninth process of dismounting and remounting only the volume in the host transitioning from the state to the operating state to remount the hard disk; A tenth step of checking whether the lock code is recorded by checking whether the state of the host is an operating host; An eleventh step of accessing the waiting host to a hard disk by recording an access enable code if the unlock code is recorded; And if the lock code is recorded, writing an access disable code to prevent the waiting host from being connected to the hard disk.

Hereinafter, with reference to the accompanying drawings will be described as follows.

1 is a block diagram illustrating a hardware configuration of a hard disk access device of a redundant host in a communication system according to an embodiment of the present invention, and FIG. 2 is a hard disk access device of a redundant host in a communication system according to an embodiment of the present invention. Figure 3 is a block diagram showing the software configuration of Figure 3 is a flow chart illustrating a hard disk access method of a redundant host in a communication system according to an embodiment of the present invention, Figure 4 is a host redundancy in the ROM processing unit in FIG. 5 is a flowchart illustrating a hard disk connection operation, FIG. 5 is a flowchart illustrating a hard disk connection operation according to host duplication at a hard disk connection unit in FIG. 3, and FIG. 6 is a host duplication at a RAM processing unit in FIG. 3. 7 is a flowchart illustrating a hard disk connection operation according to the present invention, and FIG. 7 is a timer process in FIG. 3. This is a flowchart showing hard disk reconnection and simultaneous connection prevention operations when a host is switched in a remote.

An apparatus for accessing a hard disk of a redundant host in a communication system according to an embodiment of the present invention can be roughly divided into a hardware configuration and a software configuration. As shown in FIG. 1, the first configuration is a redundant host. 10 and the second host 20, and RDC (RAID Disks Controller) (30).

The first host 10 and the second host 20 may include nonvolatile RAMs (NVRAMs) 11 and 21, dual port memories (DPRAMs) 12 and 22, and small computer system interface units (SCSIs) 13 and 23. And the CPU 14, respectively. Herein, the NVRAMs 11 and 21 perform a lock function to prevent a waiting host among the hosts 10 and 20 from accessing a hard disk. The DPRAMs 12 and 22 transfer redundancy information and hard disk access enable and disable information between a running host and a standby host among the hosts 10 and 20 using a data bus D-Bus. Do this. The SCSI units 13 and 23 are connected on the same host SCSI bus and have independent SCSI identification numbers according to the mounting positions of the hosts 10 and 20. In the case of a host, the identification number is a first SCSI identification number (SCSI ID7) and in the case of a waiting host, the identification number has a second SCSI identification number (SCSI ID6). The CPU 14 controls and processes all the above-described operations.

In addition, the RDC 30 exchanges data with the hosts 10 and 20 through a host SCSI bus and data with a hard disk through a target SCSI bus. The data mirroring and data striping are performed. It performs the function of redundancy of the hard disk through striping. Here, the host SCSI bus and the target SCSI bus are operated in an Asyncronous manner in a SCSI 2 scheme, and the transmit rate is 5 (MByte) per second, and is composed of an 8 (bit) data bus and 1 parity bit.

Secondly, as shown in FIG. 2 in the software configuration, the hard disk connection portion 40 which controls the connection between the host 10 and the hard disk in the software configuration, and the host in the ROM. A ROM processing unit 50 for controlling a redundant connection between the hard disks 10 and 20, a RAM processing unit 60 for controlling a redundant connection between the host 10 and 20 and a hard disk in a RAM, and an operating system ( 10 (msec) Period Task Module used as Real Time Clock (RTC) of Operating System) to control hard disk reconnection and simultaneous connection when the host 10 or 20 is switched It comprises a processing unit 70.

A method of accessing a hard disk of a redundant host in a communication system according to an embodiment of the present invention will be described as follows.

The ROM is divided into a ROM processing unit 50 and a RAM processing unit 60 because the ROM is booted after loading a boot file in the hard disk into RAM after booting the ROM. That is, as shown in the flowchart of FIG. 3, the ROM processing unit 50 performs a ROM processing operation (step S1), and then the RAM processing unit 60 performs a RAM processing operation (step S2). At the same time, the timer processing unit 70 performs an operation of preventing hard disk reconnection and simultaneous connection when the hosts 10 and 20 are switched (step S3).

First, as shown in the flowchart of FIG. 4, the ROM processing operation of the first step S1 in the ROM processing unit 50 first checks whether the host is in the operating state. For convenience of explanation, hereinafter, it is assumed that the first host 10 is a running host and the second host 20 is a waiting host.

However, the two redundant hosts 10 and 20 are processed from the ROM boot at the time of warm and cold start. When both the first host 10 and the second host 20 attempt to execute the hard disk connector 40, a busy state of the host SCSI bus is generated, so that the hard disk connector 40 is executed. It is checked whether its state is the first host 10 or the second host 20 (step T1).

At this time, if its state is the first host 10 in operation, the ROM processing unit 50 writes a lock code to the NVRAM 11 (step T2), thereby recording the lock code. The timer processor 70 performs an operation of preventing the hard disk connection of the second host 20 that is waiting, which is the third step S3 (step T3).

On the other hand, if the state of the second host 20 is waiting, the ROM processing unit 50 checks whether the DPRAM code recorded in the timer processing unit 70 is accessible (Access Enable) (step T4). .

Accordingly, if the DPRAM code is not accessible, the controller continuously checks whether its state is the first host 10 or the second host 20, and if the DPRAM code is access enabled or the timer processor 70 If the operation of preventing the hard disk connection of the second host 20 that is waiting for the third step (S3) in the step to perform a hard disk connection operation in the hard disk connection unit 40 (step T5) .

Secondly, as shown in the flowchart of FIG. 5, the hard disk connecting operation in the hard disk connecting unit 40 is as follows.

Since the redundant hosts 10 and 20 exist on the same host SCSI bus, they have independent SCSI identification numbers according to the mounting positions of the hosts 10 and 20, that is, the hosts 10 and 20 are mounted. The SCSI identification numbers are distinguished according to the master and slave positions and connected to the target SCSI bus (step T6).

At this time, when both the hosts 10 and 20 reset the SCSI identification number, since the first host 10 is affected by the bus reset of the second host 20, its own state is operated. Reset the SCSI bus only when it is busy (step T7).

The hard disk connector 70 mounts the disk driver (step T8), mounts the disk buffer cache (step T9), and then mounts the volume (File Manager) (step T10).

Third, as illustrated in the flowchart of FIG. 6, the RAM processing operation of the RAM processing unit 60 may be performed to perform the ROM processing operation of the first step S1 of the ROM processing unit 50. (Step T11).

Then, the RAM processing unit 60 checks whether its own state is a running host or a waiting host (step T12). At this time, if its own state is a waiting host, it terminates as it is and its state is operating. If the host, the unlock code (Unlock Code) is recorded in the NVRAM (11, 21) (step T13), and by recording the corresponding unlock code to perform the reconnection operation of the host waiting in the timer processing section 70 to the hard disk. (Step T14).

Fourth, as shown in the flowchart of FIG. 7, a host and a hard disk are waiting for the redundancy operation of the hosts 10 and 20 in the timer processing unit 70 and the switching of the hosts 10 and 20. The reconnection operation of the disk and the simultaneous connection prevention operation of the hard disk are as follows.

When the timer processing unit 70 performs a redundancy operation of the hosts 10 and 20 and the transfer occurs from the host to the waiting host, the timer processor 70 determines the state transition of the host 10 and 20 (step T15). The host that transitions from the standby host to an operational state first remounts the hard disk (step T16).

However, in the related art, the dismounting process of the hard disk proceeds in the reverse order of the hard disk access operation in the hard disk access unit 40. That is, after dismounting the volume and dismounting the disk buffer cache, the disk is mounted. The driver was dismounted.

However, the remounting in the timer processing unit 70 may shorten the time for the standby host to access the hard disk by dismounting and remounting only the volume.

Then, it is checked whether its own state is an operating host or a waiting host (step T17). At this time, if the own state is an active host, the code recorded in the NVRAM 11 or 21 is a lock code or unlocked. Check if it is a code (step T18).

When the code recorded in the NVRAM 11 and 21 is the unlock code, the access enable code is recorded in the DPRAM 12 and 22 (step T19), and the waiting host writes the corresponding access enable code. The connection operation with the hard disk is performed (step T20).

On the other hand, if the code recorded in the NVRAM 11, 21 is a lock code, an access disable code is recorded in the DPRAM 12, 22 (step T21), and the host waiting for recording the corresponding access enable code is waiting. The hard disk connection prevention operation is performed (step T22).

As described above, when the DAD hard disk is connected to and used by the present invention, the host is transferred by remounting the hard disk of the host which has been switched from the standby state to its own state and the relative state periodically. Reconnect the hard disk at the same time, prevent access to the hard disk simultaneously by using NVRAM lock and DPRAM access code, and connect the hard disk by resetting the host SCSI bus according to the host's SCSI identification number classification and host status. The hard disk is accessed from ROM according to the host status, NVRAM lock code, and DPRAM access code, and the hard disk is accessed from RAM by writing the unlock code to VNRAM according to the host status. By reconnecting the hard disk, the communication system can We can proceed with the service it is possible to improve the reliability and efficiency of the communication system.

Claims (5)

After performing the ROM processing, the RAM processing may be performed to perform a connection operation between the host and the hard disk in operation, and at the same time, a reconnection and prevention of simultaneous access between the waiting host and the hard disk may be performed. Hard disk access method for redundant hosts in a communication system. The method of claim 1, wherein the ROM processing comprises: a first process of checking whether a state of the ROM is a host in operation; A second step of recording a lock code and performing a connection prevention operation between the waiting host and a hard disk if its status is the active host; A third step of checking whether a DPRAM code is access enabled if its state is not the operating host; And a fourth process of performing a connection preventing operation between the standby host and the hard disk or performing a connection operation between the operating host and the hard disk when the DPRAM code is accessed. How to attach a hard disk on my redundant host. The method of claim 1, wherein the RAM processing operation comprises: a fifth process of confirming whether a state of the host is the operating host after the ROM processing operation is performed; And a sixth step of recording an unlock code and performing a reconnection operation of the hard disk with the waiting host if its status is the active host. The method of claim 1, wherein the connection between the host and the hard disk in operation is performed by classifying the SCSI identification number according to the mounted master and slave positions and connecting the target to the target. A seventh step of resetting the bus; And an eighth step of mounting the disk driver, the disk buffer cache, and the volumes in order, respectively. 4. The method of claim 1, 2, or 3, wherein the re-connecting and preventing simultaneous access of the standby host and the hard disk are performed only by disabling the volume in the host that detects the state transition of the host and transitions from the standby state to the operating state. A ninth step of remounting the hard disk by mounting and remounting; A tenth step of checking whether the lock code is recorded by checking whether the state of the host is an operating host; An eleventh step of accessing the waiting host to a hard disk by recording an access enable code if the unlock code is recorded; And a twelfth step of preventing the standby host from being connected to the hard disk by recording an access disable code if the lock code is recorded.

Images