CN100456253C - Method for protecting cache data of storage system - Google Patents

Abstract

A method for protecting cache data of a storage system is used for protecting the cache data of the storage system with a battery power supply function. The invention reserves a reserved area in the system random access memory for recording the information of the disk cache description item. When the storage system is powered off, the data stored in the random access memory of the system is protected through the battery power supply function. When the storage system is restarted, the data stored in the disk cache in the random access memory of the system is reserved through the information stored in the reserved area. And after the storage system is restarted, the data stored in the disk cache in the system random access memory is written into the corresponding block device, so that the power-off protection function of the data stored in the disk cache in the system random access memory is realized.

Description

Method for protecting cache data of storage system

technical field

The invention relates to a data protection method, in particular to a data protection method of a cache of a storage system.

Background technique

In a computer system, the calculation speed of a central processing unit (CPU) is extremely fast, but it takes a long time to directly read and write block devices (such as hard disks, SCSI hard disks, disk arrays, etc.), especially for block devices. At first, the CPU uses the PIO (Programmed I/O) method to read and write block devices, that is, the CPU must wait for the completion of the read and write operations on the block device before continuing to perform subsequent tasks, which greatly wastes CPU time and reduces CPU resources. In order to solve this problem, an interrupt method is adopted, that is, when the CPU reads and writes the block device, it can perform other tasks. tasks, which improves the efficiency of CPU usage. However, this method does not make full use of the characteristics of random access memory (Random Access Memory, RAM) with a faster read and write speed. Therefore, a disk cache (Disk Cache) method is adopted to further improve system performance, that is, a part of system random access memory is used as a disk cache for block device read and write operations. When the device is written, the data to be written, the device name and the location of the device are recorded in the disk cache, and the CPU considers that the write operation is over, and then the special task of the system to read and write the device periodically or When the random access memory resources of the system are insufficient, the write operation stored in the disk cache is completed, but the data in the disk cache continues to be retained; when the device is read, it is first determined whether the disk cache needs to be read The data, if any, are read directly. Otherwise, the CPU issues a task to read the device, the current task will be reserved, and the CPU can perform other tasks. When the data is read from the device, the read data will be added to the disk cache and an interrupt signal will be sent. to the CPU, the CPU can continue to perform read operations on the device. In order to improve the read and write efficiency of the block device, the driver of the block device (Driver) adopts some methods to improve the read and write efficiency, such as elevator algorithm, pre-read algorithm, delayed write algorithm, etc.

Each description item of the disk cache description item (Disk Cache Item) in the operating system represents a physical memory page (Page) or block (Block). The management of disk cache generally adopts the following method: disk cache is generally stored in three least recently used (Least Recently Use, LRU) link serials of dirty (dirty), clean (clean) and locked (locked). Among them, the dirty link sequence contains all the disk cache description items that need to be written to the block device, the lock chain sequence contains all the disk cache description items that are being written to the block device, and the clean link sequence contains the latest stored in Block device data in disk cache. Therefore, disk cache descriptors can be migrated serially between these three links. When the CPU executes the write operation to the block device, it will first check whether there is a corresponding disk cache description item from the clean link sequence, and if so, delete the corresponding disk cache description item from the clean link sequence. Change the data in the disk cache description item and add this disk cache description item to the dirty link sequence; if not, allocate the disk cache description item from the system random access memory, modify the data in the disk cache description item and Add this disk cache descriptor entry to the dirty link serialization. Then, at a certain time interval, when the system random access memory resources are insufficient, when the random access memory occupied by the disk cache reaches the upper limit set by the user, or the user forcibly clears the disk cache, the CPU will call the block device write operation The task of adding part of the disk cache description items in the dirty link sequence to the lock link sequence to write to the block device, adding the written disk cache description items to the clean link sequence, and deleting the clean link sequence The least-recently-used disk cache descriptors in the cache, deleted disk cache descriptors are reclaimed to the system's free random access memory for system use. In storage-centric systems, large amounts of random-access memory are used as disk caches.

However, the process of adding disk cache descriptors to the dirty link queue and adding disk cache descriptors written to the block device to the clean link queue is asynchronous, and often adds disk cache descriptors to the dirty link queue The process is faster. Therefore, the data stored in the disk cache is not written to the real block device. If the system loses power, the data in the disk cache is lost. Therefore, in some storage products (eg, SAN, NAS, disk array) that have high requirements on data integrity and security, this part of data needs to be protected and processed. Many vendors offer battery-backed RAM solutions.

Common battery-powered solutions include RAID Adapter, Smart Array, and battery-backed RAM adapters. It is a device of the computer system, which provides its own device driver, which can speed up the speed of writing data and the data throughput. It transfers the data from the disk cache in the random access memory of the computer system to its own internal buffer, and then processes the data through the device itself. Among them, RAID Adapter and Smart Array write the data in their own buffers to the storage device connected to themselves, and the battery-powered random access memory adapter r only stores data. Because of its battery power supply function, it can protect the data in its internal buffer from being lost within a period of time after the system is powered off, and the data in its own buffer can be written into the storage device after the system returns to normal.

However, this method of the known technology can only protect the data transferred from the random access memory of the storage system to its internal buffer when the system is powered off, and cannot protect the data stored in the disk cache in the random access memory of the computer system.

Contents of the invention

In order to solve the problems and defects in the above-mentioned known technologies, the purpose of the present invention is to provide a method for protecting cached data of a storage system, which is used to protect a battery-backed (Battery Backed) function when the computer system is powered off. The cached data of the storage system.

A method for protecting cached data of a storage system provided by the present invention includes the following steps: when the storage system is started, a reserved area of an appropriate length is reserved at an appropriate start position of the random access memory of the storage system; After startup, monitor the operation of the dirty link serial and clean link serial; when the system adds a disk cache description item to the dirty link serial, calculate the reserved area according to the location of the disk cache description item in the random access memory The serial number of an entry (Entry); store the location of this disk cache description item in the random access memory, the size of the random access memory of the disk cache description item, the name of the block device written by the disk cache description item, and the disk The location information of the cache descriptor in the block device is entered here; when the system adds a disk cache descriptor to the clean link sequence, an entry for the reserved area is calculated according to the location of the disk cache descriptor in random access memory Clear the information stored in this entry; when the storage system is powered off, use the battery power supply function to provide power to the random access memory; when the storage system restarts, read the disk cache stored by multiple entries in the reserved area The position and size information of the description item, according to the read position and size information, retain the data stored in the random access memory at the corresponding position; when the storage system is restarted, read the disk cache description items stored in multiple entries in the reserved area According to the read information, write the data stored in the random access memory at the corresponding location to the corresponding block device; and release the disk cache space in the random access memory reserved when the storage system starts, and set the reserved area The value of each entry of is zero.

Wherein, the appropriate start position of the reserved area is the random access memory location required for the initialization of the operating system of the storage system and the location other than the location of the hardware in the random access memory map, and the appropriate length is calculated by the following formula: ((random access memory Take memory size)/(cache description item size)+1)*(reserved area entry size)+(UUID size), UUID is Universal Unique IDentifier. The entry sequence number of the reserved area is calculated by the following formula: (position of the cache description item in the random access memory)/(cache description item size).

The method for protecting the cached data of the storage system according to the present invention, wherein the reserved area is reserved by adding the start position and length information of the reserved area to the memory access established by the operating system of the storage system to start the operating system itself It is implemented in the management mechanism of the region.

The method for protecting cached data of a storage system according to the present invention, wherein the information of the disk cache description item includes: the location of the disk cache description item in random access memory and the size information of the disk cache description item .

In the method for protecting cached data of a storage system according to the present invention, the entry sequence number is calculated by the following formula: (position of the disk cache description item in the random access memory)/(disk cache description item size).

To sum up, the advantage of the present invention is that the battery power supply function of the computer system is used to realize the power-off protection function for the data stored in the disk cache in the random access memory of the computer system.

Description of drawings

Fig. 1 is a flow chart of the method of the present invention.

Wherein, the reference signs are explained as follows:

S100 When the storage system is started, reserve a reserved area of an appropriate length at an appropriate start position of the random access memory of the storage system

S110, after the storage system is started, monitor operations on dirty link serials and clean link serials

S120 When adding a disk cache description item to the dirty link sequence, calculate the sequence number of an entry of the reserved area according to the position of the disk cache description item in the random access memory

S130 stores the location of the disk cache description item in the random access memory and the size information of the disk cache description item to this entry

S140 When adding a cache description item to the clean link sequence, calculate the sequence number of an entry in the reserved area according to the position of the cache description item in the random access memory, and clear the information stored in this entry

When the storage system is powered off, the S150 provides power to the random access memory through the battery power supply function

S160 When the storage system restarts, read the location and size information of the cache description items stored in the multiple entries of the reserved area, and reserve the data stored in the random access memory at the corresponding location according to the read location and size information.

S170 After the storage system is restarted, read the information of the cache description items stored in the multiple entries of the reserved area, and write the data stored in the random access memory at the corresponding location to the corresponding block device according to the read information.

S180 releases the space of the reserved random access memory, and sets the value of each entry of the reserved area to zero

Detailed ways

Hereinafter, preferred embodiments of the present invention will be described in detail in conjunction with the accompanying drawings.

Please refer to FIG. 1, which is a flow chart of a method for protecting cached data of a storage system according to the present invention. As shown in the figure, the method includes the following steps:

When the storage system is started, a reserved area of appropriate length is reserved at an appropriate start position of the random access memory of the storage system (S100), wherein the appropriate start position of the reserved area is the random access required for the initialization of the operating system of the storage system The memory location and the location of the hardware outside the random access memory map, and the appropriate length is calculated by the following formula: ((random access memory size)/(cache description item size)+1)*(reserved area entry size)+(UUID size), UUID is a universally unique identifier. When the operating system of the storage system establishes a management mechanism for starting the memory access area required by the operating system, the start position and length data of the reserved area are added to achieve the purpose of reserving the reserved area.

After the storage system is started, the monitoring system operates on dirty link sequences and clean link sequences (S110). When it is detected that the system adds a disk cache description item to the dirty link serial, according to the formula: (the position of the cache description item in the random access memory)/(the size of the cache description item), use the disk cache description item in The position of the random access memory calculates the sequence number of an entry of the reserved area (S120), and stores the position of the disk cache description item in the random access memory and the random access memory of the disk cache description item according to the calculated entry sequence number The size information comes here (S130). When it is detected that the system adds a disk cache description item to the clean link sequence, according to the formula: (position of the cache description item in random access memory)/(cache description item size), use the disk cache description item in The location of the random access memory calculates the serial number of an entry in the reserved area, and clears the information stored in the entry according to the calculated serial number of the entry (S140).

When the storage system is powered off, the battery power supply function is used to provide power to the random access memory of the system, so as to protect the data stored in the random access memory of the system. And provide short-term power for the CPU of the storage system, so that the CPU can complete the task of synchronizing the cache data of the CPU to the random access memory of the system ( S150 ).

When the storage system is restarted, all data stored in the random access memory of the system is protected by modifying the parameter setting of the basic input/output system (BIOS) of the storage system about the self-test of the random access memory. Repeat the reservation process of S100. And read the position and the size information of the disk cache description item stored in a plurality of entries of the S100 reserved area of the system random access memory, keep the data stored in the random access memory of the corresponding position according to the read position and size information (S160 ).

After the storage system restarts, read the information of the disk cache description item stored in each entry of the reserved area, find the random access memory at the corresponding location according to the read information, and read from the random access memory at the corresponding location The data to be written in the disk cache description item, the name of the block device to be written in the disk cache description item, and the location of the disk cache description item in the block device. And write the data to be written into the corresponding block device according to the read information, so as to realize the protection of the data stored in the disk cache in the random access memory of the system (S170).

After all the data stored in the disk cache in the random access memory of the system is written into the corresponding block device, release the disk cache space in the random access memory reserved when the storage system starts, and set the random access memory The value of each entry in the reserved area is zero (S180).

Although the present invention is disclosed above with the aforementioned preferred embodiments, they are not intended to limit the present invention. Those skilled in the art should realize that without departing from the scope and spirit of the present invention disclosed in the appended claims of the present invention, changes and modifications made are within the scope of patent protection of the present invention. For the scope of protection defined by the present invention, please refer to the appended claims.

Claims (8)

1. A method for protecting the cached data of a storage system, used to protect the cached data of a storage system with a battery-powered function, the method comprising the following steps: When the storage system starts up, reserve a reserved area of a length at a start location of the random access memory of the storage system; When the storage system is started, monitor the operations on the dirty link serial and the clean link serial, and store the corresponding disk cache description item in the reserved area when monitoring adds the disk cache description item to the dirty link serial information, when it is monitored that the disk cache description item is added to the clean link serial, delete the corresponding disk cache description item information in the reserved area; When the storage system is powered off, the battery power supply function is used to provide power to the random access memory to protect the data stored in the random access memory; When the storage system restarts, read the information of the disk cache description item stored in the reserved area, and reserve the data stored in the corresponding random access memory according to the read information; and After the storage system restarts, it reads the information of the disk cache description item stored in the reserved area, and writes the data stored in the corresponding random access memory into the corresponding block device according to the read information. 2. The method for protecting the cached data of the storage system according to claim 1, further comprising the following steps: releasing the disk cache space in the random access memory reserved when the storage system is started, and setting the reserved area Multiple entries have a value of zero. 3. The method for protecting the cached data of the storage system according to claim 1, wherein the starting position is the random access memory position required for the initialization of the operating system of the storage system and the position mapped by the hardware in the random access memory outside the location. 4. The data protection method of the high-speed cache of the storage system as claimed in claim 1, wherein the length is calculated by the following formula: ((random access memory size)/(disk cache description item size)+1) *(entry size of the reserved area)+(UUID size), where UUID is a universally unique identification code. 5. The method for protecting the cached data of the storage system according to claim 1, wherein the reserved area is reserved by adding the start position and length information of the reserved area to the boot operating system established by the operating system of the storage system It is implemented in the management mechanism of the required memory access area. 6. The method for protecting cached data of a storage system according to claim 1, wherein the information of the disk cache description item includes: the location of the disk cache description item in random access memory and the disk cache description item Item size information. 7. The method for protecting the cached data of the storage system according to claim 1, wherein according to the monitoring operation, the step of storing or deleting the information of the corresponding description item of the disk cache in the reserved area further comprises the following step: When it is monitored that the disk cache description item is added to the dirty link sequence, the sequence number of the entry of the reserved area is calculated according to the position of the disk cache description item in the random access memory; storing the location of the disk cache descriptor in random access memory and the size information of the disk cache descriptor in the entry; and When it is detected that the disk cache description item is added to the clean link sequence, the sequence number of the entry of the reserved area is calculated according to the position of the disk cache description item in the random access memory, and the information stored in the entry is cleared. 8. The data protection method of the high-speed cache of the storage system as claimed in claim 7, wherein the entry sequence number is calculated by the following formula: (position of the disk cache description item in the random access memory)/(disk cache description item size).

Images