CN1641610A - Hard disk replacement control management method for network storage system - Google Patents

Abstract

A hard disk replacement control management method for network storage system, which uses the record of hard disk label data as the basis of hard disk plug management to facilitate the replacement of hard disk, firstly records the label data capable of identifying hard disk in hard disk, and records the label data in the information file of the network storage system, and detects the actions of system start and hard disk replacement, and reads the label data of hard disk and compares with the corresponding data in the system file to determine the usability of hard disk.

Description

Hard disk replacement control management method for network storage system

technical field

The invention relates to a hard disk control and management method in a data storage system, in particular to a hard disk replacement control and management method for a network storage system.

Background technique

At present, there are three main modes of disk storage on the market: CAS content addressable storage, NAS network attached storage or network attached storage, and SAN storage local area network. The latter two are the main choices in the network environment.

After two rapid developments in computer computing power and network connectivity, people's demand for the network is no longer satisfied with connectivity, but more powerful information management capabilities. As more and more critical information is converted into digital form and stored in manageable media, new demands are placed on the network's ability to store and manage information. It can be said that whether the network has high efficiency depends on its data storage and management capabilities. Today, when network storage determines the network structure, the IT industry has entered the era of storage as the core from PC and network.

At present, many NAS systems support the design of drawer-type hard disk boxes to facilitate the replacement and insertion of hard disks. The advanced system supports Hot Plug (HotPlug) function, that is, the hard disk can be replaced while the system is running. Simpler systems only support hard disk replacement after shutdown, that is, cold swap.

Frequent hard disk replacement brings difficulties in hard disk identification. In the LINUX environment, when a hard disk with SCSI interface is inserted or removed, the identification device name of other online hard disks will change, resulting in a change in the access path. This further increases the difficulty of hard disk identification.

In order to solve the above problems, different solutions have been introduced before and after, but they have not completely solved the problem. For example, some NAS operating systems developed based on Linux, such as Guardgen OS, support hot swapping of hard disks. Before plugging and unplugging the hard disk, you must use the hard disk removal function in the software to remove the corresponding hard disk logical information from the system, and then manually pull out the hard disk. In this way, the system can record each insertion and removal of the hard disk, and then the system will perform corresponding processing. But the disadvantage of this is that it does not support hard disk cold swap, that is, if the hard disk is replaced after shutdown, the system cannot correctly identify and handle it.

There are also some technologies that support hard disk cold plugging, which record the partition structure data of each hard disk in the system, and record the serial number of the file system of each partition at the same time. When the system starts to detect these information, if there is a change, it will be considered that the hard disk has been replaced, and then the system will deal with it accordingly to adapt to the change of the number of hard disks and the name of the SCSI hard disk identification device. Because this solution is designed for desktop systems, and all hard disk partitions of the same capacity on the NAS system have the same structure, and the file system is not directly created on the hard disk partition, so the serial number of the file system cannot be obtained, so this solution does not work. It is suitable as a method for identifying hard drives in NAS systems.

Therefore, how to effectively control and manage the replacement of the hard disk in the network storage system, and at the same time support the hot and cold plug operation of the hard disk, has become an urgent problem to be solved.

Contents of the invention

In order to solve the above problems, the present invention provides a hard disk replacement control and management method for a network storage system, which effectively controls and manages hard disk replacement in the network storage system, supports hot and cold plugging operations of hard disks, and can maintain hard disk unique access path.

The invention provides a hard disk replacement control management method for a network storage system. The recording of hard disk label data is used as a basis for hard disk plug-in management to facilitate hard disk replacement. The method first records a mark that can identify the hard disk in the hard disk data, and correspondingly record the tag data in the information file in the network storage system, and at the same time detect the actions of system startup and hard disk replacement, after detecting these actions, read the tag data of the hard disk and compare it with the corresponding data in the system file Compare to determine the availability of the hard disk.

The present invention identifies the usability of the hard disk by recording mark data in the implicit sector of the hard disk, and supports hot and cold plugging of the hard disk at the same time, and through these data, the access path of the hard disk can be ensured when the hard disk is plugged and unplugged. change.

Description of drawings

FIG. 1 is a flowchart of a hard disk replacement control management method for a network storage system according to the present invention;

FIG. 2 is a schematic diagram of storage locations of hard disk tag data provided by the present invention;

FIG. 3 is a schematic diagram of an initialization process of an unavailable hard disk provided by the present invention;

Fig. 4 is a schematic diagram of the processing flow of the cold plug operation provided by the present invention;

FIG. 5 is a schematic diagram of a hot swap operation processing flow provided by the present invention; and

6A to 6C are schematic diagrams of processing the unique access path of the hard disk provided by the present invention.

Wherein, the reference signs are explained as follows:

Step 101 records the tag data that can identify the hard disk in the hard disk

Step 102 correspondingly records the mark data in the information file in the network storage system

Step 103 detects the actions of system startup and hard disk replacement

Step 104 reads the marking data of the hard disk and compares it with the corresponding data in the system file to determine the availability of the hard disk

Step 301 deletes all data in the hard disk and repartitions according to system requirements

Step 302 reads the hardware information of the hard disk and generates hard disk label data

Step 303 Write the hard disk mark data into the hidden sector of the hard disk and the system information file

Step 304 records the hard disk as an available state in the system file

Step 305 establishes the link between the unique access path of the hard disk and the identification device name of the hard disk

Step 401: the system reads the tag data recorded in the hidden sector of the hard disk, and checks the validity of the tag data

Step 402 compares the data with the hard disk tag data recorded in the storage information file

Step 403 judges whether the tag data comparison is the same

In step 404, if the comparison results are the same, it is considered that the hard disk is the hard disk before the system shuts down, and no cold swap changes have taken place, and it is a hard disk available to the system

Step 405 otherwise considers that the hard disk has changed, and the hard disk is unavailable for the system

Step 501 The system detects the information that the hard disk is pulled out

Step 502 system deletes the corresponding hard disk mark data information in the storage information file

Step 503 When the user reinserts the hard disk at the position where the hard disk was pulled out

Step 504 identifies the hard disk as unavailable

Detailed ways

As shown in Figure 1, it is the flow chart of the hard disk replacement control management method for the network storage system of the present invention, first record the tag data (step 101) that can identify the hard disk in the hard disk; in the information file in the network storage system Correspondingly record this mark data (step 102); Detect system startup and the action of hard disk replacement simultaneously (step 103); Read the mark data of hard disk and compare with the corresponding data in the system file, to determine the usability of hard disk (step 104) .

Wherein the hard disk mark data is recorded in the unused hidden sector of the hard disk, and the hidden sector refers to n unused sectors of the system whose sector number is smaller than the number of magnetic heads behind the master boot sector (MBR) of the hard disk. The number of magnetic heads of the current hard disk is greater than 16, so at least 15 sectors can be used as storage locations for hard disk label data. Therefore, as shown in FIG. 2 , the first sector after the master boot sector of the hard disk can be used as the hard disk mark data storage location. At the same time, the tag data of each hard disk can be saved in the storage information file of the NAS system at the same time, and the system can identify whether the hard disk and its location have changed by comparing the contents of the two hard disk tag data. Among them, when the action of system startup or the action of inserting the hard disk is detected during the operation of the system, if the read hard disk tag data is the same as the record in the system file, it will be identified as usable; if the read hard disk If the flag data differs from the record in the system file, it is recognized as unavailable.

If the action of pulling out the hard disk is detected during system operation, the tag data of the pulled out hard disk is deleted from the system file.

The system information file also records the unique access path of each hard disk, and establishes a link with the hard disk device name. When the action of system startup or the action of inserting the hard disk is detected during system operation, a link between the unique access path of the hard disk and the name of the hard disk device is established according to the hard disk tag data; When the operation of the hard disk is disabled, the link between the unique access path of the hard disk and the device name of the hard disk is deleted.

If this hard disk is identified as being unavailable, then the hard disk initialization function of the system needs to be invoked, as shown in Figure 3, first delete all data in the hard disk and repartition according to system requirements (step 301); then read the hardware information of the hard disk And generate hard disk marking data (step 302); This hard disk marking data is written in hard disk hidden sector and system information file (step 303); This hard disk is recorded as usable state (step 304) in this system file; Finally Establish a link between the unique access path of the hard disk and the identification device name of the hard disk (step 305).

The hard disk mark data can record the following system information and hard disk information:

NAS serial number: The unique identification number of the system set by the NAS factory. The serial number of each NAS will not be repeated. The system can identify different NAS by reading this identification;

Hard disk serial number: the unique identification number set when the hard disk leaves the factory, due to length limitations, this number may be repeated;

Hard disk model: The model code set when the hard disk leaves the factory. The hard disk of the same model has the same code. The hard disk serial number and hard disk model are not used as the unique identification information of the hard disk, but normal reading of this information indicates that the hard disk controller and system BIOS are basically Normal, otherwise there may be a hardware failure in the system;

Hard disk Universal Unique Identifier (UUID): It is to solve the unique identification problem of the hard disk. The hard disk UUID is a random number composed of 128 letters and numbers generated by software. This number is generated by a special algorithm and can be Ensure that the UUID of any hard disk of any NAS is not repeated;

Hard disk installation location: IDE hard disk records IDE interface number and hard disk master-slave position, SCSI hard disk records hard disk interface information such as Channel, ID, LUN, etc., and these information will change when the hard disk is plugged in or removed or the installation location changes.

For the identification process of usability after the hard disk cold plug operation, as shown in Figure 4, when the system is restarted, the system reads the mark data recorded in the hidden sector of the hard disk, and checks the validity of the mark data (step 401); After the verification is passed, the data is compared with the hard disk tag data recorded in the storage information file (step 402); whether the judgment tag data comparison result is the same (step 403); if the comparison result is the same, then it is considered that the hard disk is before the system shutdown If there is no cold swap change, it is a hard disk available to the system (step 404); otherwise, it is considered that the hard disk has changed and it is a hard disk that is not available to the system (step 405), and the hard disk initialization function of the NAS system must be used to initialize the hard disk before continuing to use it.

For the identification process of availability after the hard disk hot plug operation, as shown in Figure 5, when the hard disk is pulled out during the system operation, at first the information that the hard disk is pulled out is detected by the system (step 501); then the system deletes the information in the storage information file The corresponding hard disk mark data information (step 502), represents that this hard disk can not continue to use; Unusable hard disk (step 504), the hard disk cannot be used, and the hard disk must be initialized with the hard disk initialization function of the NAS system before it can be used again.

Hard disk initialization refers to the process that the user invokes the hard disk initialization function of the system to mark the hard disk as usable when the system considers the hard disk unusable after replacing the hard disk. During this process, the system first deletes the original data of the hard disk, repartitions according to the requirements of the NAS system, and then reads hardware information such as the NAS serial number, hard disk model, and hard disk serial number, and then calls a special algorithm to generate the hard disk UUID, and combines these information into Hard disk label data. Write hard disk hidden sectors and storage information files. In the storage information file, the hard disk record is rewritten into an available state, and finally the link between the unique access path of the hard disk and the identification device name of the hard disk is established. In this way, the NAS system starts to use this hard disk.

Because the hard disk identification device names of the SCSI hard disks of the NAS under the Linux system are arranged sequentially, when adding or reducing hard disks, the hard disk identification device names will be rearranged. If no processing is performed, the access path of the hard disk will change accordingly. Therefore, the present invention will establish a unique access path for each hard disk in the system, which is different from the hard disk identification device name. When the hard disk does not exist, this access path cannot be used, but it will not be occupied by other hard disks. If the hard disk exists, establish a link between the unique access path of the hard disk and the identification device name of the hard disk, so that the effect of the NAS system accessing the unique access path of the hard disk and the identification device name of the hard disk is the same. When the name of the SCSI hard disk identification device changes due to the increase or decrease of hard disks, the system re-establishes the link between the unique access path of the hard disk and the identification device name of the hard disk to ensure that the access path of the hard disk remains unchanged.

As shown in FIG. 6A , assuming that the system can support 4 SCSI hard disks, the Linux system arranges the device names of the existing hard disks in order: /dev/sda; /dev/sdb; /dev/sdc; /dev/sdd. The present invention sets up a unique access path (it is different from equipment name) for each hard disk: /dev/shda; HDD 1 /dev/shda /dev/sda HDD 2 /dev/shdb /dev/sdb HDD 3 /dev/shdc /dev/sdc HDD 4 /dev/shdd /dev/sdd

When the system starts, if the hard disk 3 is not inserted, as shown in Figure 6B, then the Linux system arranges the device names of the existing hard disks in sequence. The installation location establishes a link between the unique access path of the hard disk and the device name of the hard disk. The corresponding relationship is: HDD 1 /dev/shda /dev/sda HDD 2 /dev/shdb /dev/sdb HDD 3 /dev/shdc null HDD 4 /dev/shdd /dev/sdc

If the hard disk 3 is inserted during system operation, as shown in FIG. 6C , the device name of the hard disk 3 is recognized by the Linux system as /dev/sdd. The present invention reads the "hard disk installation location" data marked on the hard disk, and establishes a link between the unique access path of the hard disk and the device name of the hard disk according to the installation location of the hard disk. The corresponding relationship is: HDD 1 /dev/shda /dev/sda HDD 2 /dev/shdb /dev/sdb

Hard disk 3 /dev/shdc /dev/sdd Hard disk 4 /dev/shdd /dev/sdc

For example, when the hard disk 2 is pulled out while the system is running, the device name /dev/sdb of the hard disk 2 cannot be accessed, resulting in the inaccessibility of the only access path /dev/shdb of the hard disk. At this time, the system does not need to handle it.

The above are only preferred embodiments of the present invention, and are not intended to limit the scope of the present invention; all equivalent changes and modifications made within the scope of the patent application for the present invention without departing from the spirit and scope of the present invention are all It should belong to the scope of protection required by the patent claims of the present invention.

Claims (9)

1, a kind of hard disk that is used for network storing system is changed the control and management method, is the foundation that plug is managed as hard disk by the record of hard disk flag data, and to make things convenient for the replacing of hard disk, wherein this method comprises:

Record can identify the flag data of hard disk in hard disk;

This flag data of corresponding record in the message file in this network storing system;

The action that detection system starts and hard disk is changed; And

Read the flag data of hard disk and compare, to determine the availability of hard disk with corresponding data in the system file.

2, the hard disk that is used for network storing system as claimed in claim 1 is changed the control and management method, and wherein this hard disk flag data is recorded in the obsolete implicit sector of hard disk.

3, the hard disk that is used for network storing system as claimed in claim 1 is changed the control and management method, and wherein this flag data comprises: the network storing system sequence number; The hard disk sequence number; The hard disk model; The hard disk universal unique identifier; Position of installing hard disk.

4, the hard disk that is used for network storing system as claimed in claim 1 is changed the control and management method, wherein detect this hard disk changes when moving to the insertion hard disk when the action that detects system start-up or in system's operational process, if the hard disk flag data that is read is identical with record in the system file, then identification its be available; If the hard disk flag data that is read is different with the record in the system file, then identification its be unavailable.

5, the hard disk that is used for network storing system as claimed in claim 1 is changed the control and management method, wherein change action when extracting hard disk, also comprise the step of from system file, deleting the flag data of the hard disk of being extracted when detecting this hard disk in system's operational process.

6, the hard disk that is used for network storing system as claimed in claim 1 is changed the control and management method, wherein also records unique access path of each hard disk in this system information file, and links with the foundation of hard disc apparatus name.

7, the hard disk that is used for network storing system as claimed in claim 1 is changed the control and management method, after wherein this determines the step of hard disk availability, if this hard disk of identification is unavailable, then also comprises the steps:

Delete in this hard disk all data and according to system requirements subregion again;

Read the hardware information of this hard disk and generate the hard disk flag data;

This hard disk flag data is write in implicit sector of hard disk and the system information file;

In this system file, be upstate with this hard disk recording; And

Set up linking between the unique access path of hard disk and the hard disk identification equipment name.

8, the hard disk that is used for network storing system as claimed in claim 1 is changed the control and management method, wherein detect this hard disk and change action when inserting hard disk, also comprise according to this hard disk flag data and set up the step that links between the unique access path of hard disk and the hard disc apparatus name when the action that detects system start-up or in system's operational process.

9, the hard disk that is used for network storing system as claimed in claim 1 is changed the control and management method, wherein change action when extracting hard disk, also comprise and delete linking between the unique access path of this hard disk and this hard disc apparatus name when detecting this hard disk in system's operational process.

Images