WO2014094303A1 - 一种监控记录管理方法与装置 - Google Patents
一种监控记录管理方法与装置 Download PDFInfo
- Publication number
- WO2014094303A1 WO2014094303A1 PCT/CN2012/087170 CN2012087170W WO2014094303A1 WO 2014094303 A1 WO2014094303 A1 WO 2014094303A1 CN 2012087170 W CN2012087170 W CN 2012087170W WO 2014094303 A1 WO2014094303 A1 WO 2014094303A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- monitoring
- record
- memory
- monitoring record
- records
- Prior art date
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 899
- 238000007726 management method Methods 0.000 title claims description 19
- 238000000034 method Methods 0.000 claims abstract description 37
- 238000004891 communication Methods 0.000 claims description 12
- 230000008859 change Effects 0.000 description 8
- 230000006870 function Effects 0.000 description 6
- 238000001514 detection method Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000001960 triggered effect Effects 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 230000009172 bursting Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000004590 computer program Methods 0.000 description 1
- 238000013523 data management Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
- G06F11/3466—Performance evaluation by tracing or monitoring
- G06F11/3476—Data logging
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/3003—Monitoring arrangements specially adapted to the computing system or computing system component being monitored
- G06F11/3037—Monitoring arrangements specially adapted to the computing system or computing system component being monitored where the computing system component is a memory, e.g. virtual memory, cache
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/40—Data acquisition and logging
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/3003—Monitoring arrangements specially adapted to the computing system or computing system component being monitored
- G06F11/3034—Monitoring arrangements specially adapted to the computing system or computing system component being monitored where the computing system component is a storage system, e.g. DASD based or network based
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
- G06F11/3409—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment for performance assessment
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
- G06F11/3442—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment for planning or managing the needed capacity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
- G06F11/3466—Performance evaluation by tracing or monitoring
- G06F11/3485—Performance evaluation by tracing or monitoring for I/O devices
Definitions
- the present invention relates to the field of storage, and in particular to a management technique for monitoring records. Background technique
- the data in the storage system has lifecycle characteristics. For example, a recently produced online video, many people see, the traffic is very high, then on the video server, this video is hot data; Over time, a few months later, this video has been less noticed. The video traffic is drastically reduced. On the video server, this video is hot data. After a few years, almost no one watched this video, and almost no one visited it. On the video server, this video is cold. Point data. Perform tiered storage, migrate hotspot data to a high-performance tier with fast read speed in the storage system, and improve the read speed of these data; migrate the cold spot data to a tiered storage system in the storage system. The low performance level meets the performance and capacity requirements of storage systems at a lower price.
- a data block is a basic storage unit in a storage system, and data is stored in a data block.
- data is stored in a data block.
- the monitoring parameters of each data block such as access frequency, 10 size, etc.
- a storage medium such as memory
- the life cycle of the data in the data block can be monitored by analyzing the monitoring data.
- the number of monitoring records is determined by the total number of data blocks, and it is often difficult to meet the actual needs of users. For example, when a large number of data blocks are stored in the storage system, a large number of monitoring records pose a huge challenge to the capacity of the storage medium.
- the present invention provides a monitoring record management method and apparatus, which can change the monitoring granularity of monitoring records.
- an embodiment of the present invention provides a monitoring record management method, in which a plurality of monitoring records are stored in a memory, and each monitoring record monitors one data block in the memory, and one monitoring record is a a set of various monitoring parameters of the data block, where the monitoring parameter includes, for example, the number of read/write times per second of the monitored data block, the method comprising: reading at least one monitoring parameter from each monitoring record in the memory; The monitoring record having one or more monitoring parameter differences meeting the preset threshold is read, and the read monitoring records are combined to generate a combined monitoring record, wherein the combined monitoring record records the identifier of the monitored data block.
- the monitoring record is written into the memory, and the combined monitoring record is used to monitor the data block monitored by each monitoring record before the merge.
- the method further includes: monitoring an I/O request for the data block in the memory; and merging the monitoring data record of the monitored data block An I/O request record for updating various monitoring parameters of the combined monitoring record in the memory.
- the method further includes: reading the merged monitoring record from the memory, and splitting the combined monitoring record into at least one split monitoring record, and each of the split monitoring records Recording is used for monitoring at least one data block, wherein the sum of the monitoring parameter values is equal to the value of the monitoring parameter in the combined monitoring record; deleting the read combined monitoring record from the memory, and Each of the split monitoring records is written to the memory.
- an embodiment of the present invention provides a method for monitoring record splitting.
- a plurality of monitoring records are stored in a memory, and each monitoring record monitors at least one data block in the memory, and each monitoring record is a monitored data block.
- a set of monitoring parameters the monitoring parameter includes, for example, IOPS per second of the monitored data block, the method comprising: reading at least one item from each monitoring record in the memory Monitoring parameters;
- each monitoring parameter is split according to the preset monitoring record, and at least 2 monitoring records are generated after the splitting, and each split monitoring record carries the data offset.
- Address and data length, in the split monitoring record the sum of the monitoring parameter values is equal to the value of the monitoring parameter in the combined monitoring record; the monitoring record after the split is written into the memory, and the monitoring record is used after the splitting Monitor the data of the offset address and data length carried.
- the values of the monitoring parameters in each monitoring record after the splitting are specifically: the values of the monitoring parameters in the monitoring record before the splitting, according to the monitoring records after the splitting
- the length of the data carried in the data is distributed to the monitoring records after the split.
- an embodiment of the present invention provides a monitoring and recording management apparatus, wherein a plurality of monitoring records are stored in a memory, and each monitoring record monitors a data block in the memory, and one monitoring record is a monitoring parameter of a data block.
- the monitoring parameter includes, for example, the number of read/write times per second of the monitored data block
- the device includes: a first reading module, configured to read at least one monitoring parameter from each monitoring record in the memory; And the monitoring record having one or more monitoring parameter differences meeting the preset threshold is read from the memory, and the read monitoring records are combined to generate a combined monitoring record, wherein the combined monitoring records record the monitored records.
- the identifier of the data block, the read monitoring record is deleted from the memory, wherein the value of each monitoring parameter in the combined monitoring record is equal to the value of each monitoring parameter when the monitored data block is regarded as a data block a first writing module, configured to write the combined monitoring record into the memory, where the combined monitoring record is used to monitor each of the pre-merger Control records the data blocks monitored.
- the first write module is further configured to monitor an I/O request for a data block in the memory, and combine the I/O request of the monitored data block by the monitoring record. Recording, used to update various monitoring parameters of the combined monitoring record in the memory.
- the first splitting module is configured to read the merged monitoring record from a memory, Splitting the combined monitoring record into at least one split monitoring record, each of the split monitoring records is used for monitoring at least one data block, wherein the value of each monitoring parameter is added to the split monitoring record The value of the monitoring parameter in the combined monitoring record; the read combined monitoring record is deleted from the memory, and each of the split monitoring records is written into the memory.
- an embodiment of the present invention provides a monitoring and recording splitting device, wherein a plurality of monitoring records are stored in a memory, and each monitoring record monitors at least one data block in the memory, and each monitoring record is a monitored data block.
- a set of monitoring parameters the monitoring parameter includes, for example, IOPS per second of the monitored data block
- the device includes: a second reading module, configured to read at least one monitoring from each monitoring record of the memory
- the second splitting module is configured to split the monitored monitoring parameters of each of the monitored monitoring parameters into a preset threshold, and generate at least two monitoring records after splitting, after each splitting
- the monitoring record carries the offset address and the data length of the data.
- the split monitoring record the sum of the monitoring parameter values is equal to the value of the monitoring parameter in the combined monitoring record; the second writing module is used for The post-split monitoring record is written into the memory, and the split monitoring record is used to monitor the data of the carried offset address and data length.
- each monitoring parameter is split according to a preset monitoring value, and at least two monitoring records are generated after splitting,
- the split monitoring record carries the offset address and the data length of the data.
- the sum of the monitoring parameter values is equal to the value of the monitoring parameter in the combined monitoring record, and the monitoring is performed before the splitting.
- the values of the monitoring parameters in the record are distributed to the monitoring records after the split according to the length of the data carried in the monitoring records after the split.
- the second reading module is specifically configured to: read random 10 from each monitoring record in the memory The number and the total number of 10, the threshold is the ratio of the random 10 number to the total 10 number.
- an embodiment of the present invention provides a storage controller, where the storage controller includes a CPU, a communication interface for communicating with the CPU, and a memory for communicating with the CPU: the communication interface is configured to communicate with the memory; There are a plurality of monitoring records stored in each monitoring record, and one monitoring record is a set of monitoring parameters of a data block, and the monitoring parameter includes the number of readings per second of the monitored data block.
- IOPS IOPS
- the processor configured to execute a program, wherein the program is configured to: read at least one monitoring parameter from each monitoring record in the memory; The monitoring record having one or more monitoring parameter differences meeting the preset threshold is read in the memory, and the read monitoring records are combined to generate a combined monitoring record, wherein the monitored monitoring record records the monitored data block. Identifying, the read monitoring record is deleted from the memory, where the value of each monitoring parameter in the combined monitoring record is equal to the value of each monitoring parameter when the monitored data block is regarded as a data block; The combined monitoring record is written into the memory, and the combined monitoring record is used to monitor the data blocks monitored by the monitoring records before the combining.
- an embodiment of the present invention provides a storage controller, where the storage controller includes a CPU, a communication interface that communicates with the CPU, and a memory that communicates with the CPU: the communication interface is configured to communicate with the storage;
- the memory is configured to store a plurality of monitoring records, each monitoring record monitors at least one data block in the memory, each monitoring record is a set of monitoring parameters of the monitored data block, and the monitoring parameter includes the monitored data.
- the number of reads and writes per second of the block is IOPS; and the processor is configured to execute a program, wherein the program is configured to: read at least one monitoring parameter from each monitoring record in the memory; and to read the monitored parameter,
- Each monitoring parameter is split according to a preset monitoring record, and at least two monitoring records are generated after splitting, and each split monitoring record carries an offset address and a data length of the data, and the splitting is performed.
- the sum of the monitoring parameter values is equal to the value of the monitoring parameter in the combined monitoring record; the post-split monitoring record is written into the memory, and the split monitoring record is used to monitor the carried offset address and data. Length data.
- the monitoring record can be split or merged, that is, the monitoring granularity of the monitoring record can be changed.
- FIG. 1 is a top view of a first embodiment of the present invention
- Figure 2 is a structural view of an embodiment of the present invention
- Figure 3 is a structural view of an embodiment of the present invention
- FIG. 4 is a flowchart of a method for monitoring records management according to an embodiment of the present invention.
- FIG. 5 is a flowchart of an embodiment of a method for monitoring record splitting according to an embodiment of the present invention
- FIG. 6 is a structural diagram of a monitoring record management apparatus according to an embodiment of the present invention.
- FIG. 7 is a structural diagram of a monitoring and recording splitting apparatus according to an embodiment of the present invention. detailed description
- a storage system is a system for providing a data storage service, and may be composed of a physically relatively independent storage controller and memory, or a server having a storage function.
- a data block (Chunk) is a basic logical storage unit in a memory for storing data.
- the data block is a logically continuous storage space in the memory, and each data block is mapped with a physical storage medium, and the physical storage medium may be discontinuous. of.
- the data block can be fixed length or variable length, and its capacity can be only a few K bytes or several G bytes.
- the stored data needs to be monitored, and the monitoring record is generated by using the data block as a basic monitoring unit, and the monitoring record is composed of at least one monitoring parameter.
- Each of the monitoring parameters is used to monitor a dimension of the read and write status of the data block.
- a monitoring record can be set for each data block, which needs to be stored using a storage medium, and the monitoring record can be stored with the business data.
- these monitoring records can also be stored separately from the business data, such as memory stored in the storage system in.
- the memory can be a random access memory (RAM) solid state disk (SSD) or flash memory (Flash). If the read/write speed is not high, you can also use storage media such as disk and tape.
- the length of the 10 request exceeds the size of the data block, then this 10 requests are split according to the size of the data block, and then stored in different data blocks.
- the monitoring record of each data block is requested by 10
- 10 requests are statistical basis. For example, the length of the write 10 request is 100K, and the single data block is only 30K, then the 10 request can be split into 4 sub-10 requests, the data is written into 4 data blocks, and the monitoring records of each data block are only Monitor the 10 sub-requests corresponding to this data block. That is to say, in the embodiment of the present invention, in the case where 10 requests for splitting, the term "10 request" refers to the split 10 request.
- FIG. 1 it is a topology diagram to which Embodiment 1 of the present invention is applied.
- the monitoring program can monitor the 10 requests sent by the service; monitor and obtain various monitoring parameters such as IOPS and 10 sending frequency of the data block, and then monitor the parameters.
- the composed monitoring record is stored in the memory; the data lifecycle management program analyzes the monitoring record, and then manages the data block in the memory by issuing a data management instruction to the memory, for example,
- the frequent data blocks are migrated to the area where the reading and writing speed is faster in the memory.
- the memory in the figure is set in the controller of the storage system, and the monitoring program and the data lifecycle management program run in the CPU in the storage controller.
- Each monitoring record includes several monitoring parameters, and the monitoring parameter records the read/write status of the data block.
- the monitoring parameters may include: Input Operations Per Second; Data block write operations per second ( Output Operations Per Second) One or more of the number of read/write operations per second (I0PS, Input/Output Operations Per Second).
- the monitoring parameters may also include one or more of a read 10 count (read 10 count ), a write 10 (write 10 count ) count, or a read/write 10 total (10 count ).
- the monitoring parameter may further include: read 10 size (read 10 size), the size of the read 10 refers to the size of a read read data requested by 10, and 10 requests that the data desired to be read may be smaller than the size of the data block, ie Read only a part of the data in a data block; write 10 size (write 10 size), the size of write 10 refers to the size of a request to read 10 data, 10 request that the data you want to write can be smaller than the data block Size, that is, only a part of a data block; number of read operations; number of write operations; randomness, randomness refers to the ratio of random 10 to a total of 10; frequency of 1/0, frequency of 1/0 is unit time Number of requests within 10; 1/0 delay (10 lantency) Size (the response time of the storage server to the host request).
- read 10 size read 10 size
- the size of the read 10 refers to the size of a read read data requested by 10
- 10 requests that the data desired to be read may be smaller than the size of the data
- the memory required for the monitoring record is also very large.
- a plurality of data blocks can be monitored by using one monitoring record to avoid the occurrence of new data blocks due to insufficient memory space. Create a monitoring record.
- the embodiment of the present invention can also monitor different ranges of the same data block by multiple monitoring records, and enhance the fineness of the monitoring result.
- Embodiment 1 is a diagrammatic representation of Embodiment 1:
- One embodiment of the present invention provides a monitoring record management method that can be applied to a storage controller.
- the storage controller stores a plurality of monitoring records in the memory, each monitoring record monitors one data block in the memory, and one monitoring record is a set of monitoring parameters of a data block, and the monitoring parameter includes, for example, the monitored data block.
- the monitor monitors the read and write requests of the data block and updates the monitoring records in the memory in real time with the monitoring results.
- NAS network storage technology
- SAN storage area network
- the storage controller is physically independent of the storage.
- the storage server In the storage server, the storage server itself has a storage controller. Equivalent to the storage controller being integrated in the storage server, so the storage server itself is implemented by this method. In other forms of storage systems, this method can be performed by a device having similar computing capabilities as the storage controller. In the embodiment of the present invention, whether or not it is a physically independent storage controller, only functions having similar functions are collectively referred to as a
- the storage system 4 is composed of a storage controller 41 and a memory 42, wherein the storage controller 41 includes a CPU 411, a communication interface 412 in communication with the CPU 411, and a memory 413 in communication with the CPU 411.
- the memory controller communicates with memory 42 via communication interface 412.
- the CPU 411 runs a computer program, which may include program code, and the running program code may perform the steps as shown in FIG.
- Step 11 Read at least one monitoring parameter from each monitoring record in the memory.
- each monitoring record is used to monitor one data block in the memory, each monitoring The record is composed of at least one monitoring parameter, and each monitoring parameter is obtained by monitoring the read and write of the data block to the 10 request.
- each monitoring record monitors one data block and monitors a total of 5 data blocks.
- Each monitoring record includes three monitoring parameters.
- monitoring record A includes monitoring parameter A1, monitoring parameter A2, and monitoring parameter A3.
- the parameters can be, for example, IOPS, read/write frequency, and 10 counts, respectively.
- Step 11 can be triggered by the user actively, or can be triggered by the storage system according to the preset threshold condition. For example, if the existing monitoring parameters have already occupied the memory, that is, the memory has no more space to store more monitoring records, go to step 11. At this time, if the new data block needs to be monitored again, there will be no case where the monitoring parameters of the new data block are not available for use. Or the monitored data blocks are generally small, for example, only a few K. It is not necessary to set the monitoring parameters separately for each data block, or you can perform 11 steps.
- Step 12 Read, from the memory, a monitoring record having one or more monitoring parameter differences that meet a preset threshold, and combine the read monitoring records to generate a combined monitoring record, where the recorded records are monitored.
- the identifier of the data block, the read monitoring record is deleted from the memory, wherein the value of each monitoring parameter in the combined monitoring record is equal to each monitoring when the monitored plurality of data blocks are regarded as one data block The value of the parameter.
- the read monitoring records are combined to generate a combined monitoring record, which is to merge the monitored monitoring records with one or more monitoring parameter differences that meet the preset threshold. .
- the threshold value of each monitoring parameter difference can be specified by the user.
- the merged monitoring record carries the identifier of the monitored data block, and the identifiers of different data blocks are different.
- each monitoring record Before the merger, each monitoring record also records the identification of the data block monitored by each.
- the combined monitoring record monitors a plurality of data blocks, so the identification of the plurality of data blocks is recorded in the combined monitoring record, and the identifiers of the different data blocks are different. Through the identification, the user can know which data is monitored by the monitoring parameter. Piece.
- the threshold value detection of the difference is performed on the content read in step 11.
- the number of monitored monitoring parameters may be one or more, that is, only one monitoring parameter may be judged. If the difference of the monitoring parameters meets the threshold in different monitoring records, a matching manner of the threshold is If the monitoring record is less than or not greater than a specific value, the monitoring records are read out as a whole.
- the monitoring parameters read in step 101 include IOPS, monitoring record 8, monitoring record ⁇ monitoring record ⁇ , monitoring record D, monitoring record
- the IOPS recorded in E are 50, 52, 60, 61, 70, respectively, and the IOPS has a threshold of 3, which means that the monitoring records with the difference of IOPS less than 3 are in compliance with the combined conditions.
- the monitoring parameters of the monitoring record A and the monitoring record B are read out and combined separately, and the combined monitoring records simultaneously record the monitoring records of the data block A and the data block B; the monitoring record C and the monitoring record D are also recorded.
- the memory will be read out and merged, and the combined monitoring record will record the monitoring records of data block C and data block D at the same time.
- the following examples are further based on the previous examples. It is assumed that in addition to the IOPS, the monitoring parameter 10 count (10 count) is compared, the monitoring parameters, Monitoring record ⁇ monitoring record, monitoring record 0, monitoring record E, 10 counts are 10, 11, 20, 30, 40, respectively, the threshold of 10 count is 5, that is, only monitoring parameter A and monitoring parameter B 10
- the monitoring record A and the monitoring record B satisfy the threshold requirements of IOPS and 10 counts at the same time.
- the monitoring record C and the monitoring record D satisfy the threshold requirement of IOPS, but do not satisfy the threshold requirement of 10 counts. Therefore, only monitor record A and monitor B will be read out and combined to generate a combined monitor record. Monitor record C and monitor record D will not be read from the memory.
- the value of the combined monitoring parameter is equal to the value of each monitoring parameter when the monitored data block is regarded as a data block, or the value of the monitoring parameter obtained when monitoring the monitored data block set, in the data block set.
- a single 10 request for any data block is recorded as a 10 request for the entire data block set.
- the value of the combined monitoring parameter is generated by combining the monitoring parameters of all monitored data blocks, and the data block monitored by the combined monitoring parameter is a set of data blocks monitored by the combined monitoring parameters. Hehe.
- the pre-merging monitoring records record the I/O request records of the respective monitored data blocks, and the combined monitoring records record the I/O request records of the monitored data block sets.
- the merged monitoring records are originally monitored with data blocks, and the values of the monitoring records in the combined monitoring parameters are equal to the values obtained by monitoring the data blocks together, and the monitoring records in the combined monitoring parameters may also be referred to as merges. Monitor parameters.
- the monitoring parameter A and the monitoring parameter B are all composed of three monitoring parameters, which are IOPS, 10 counts, 10 frequencies, wherein the value of the IOPS in the monitoring record A is 50, 10 The value of the count is 10, and the frequency of 10 is 10; the value of IOPS in monitoring record B is 52, the value of 10 count is 11, and the frequency of 10 is 20.
- the values of the monitoring parameters are directly added, because if the set consisting of the data block A and the data block B is regarded as a new data block AB, the IOPS is taken as an example, and the data block A or the data block is used.
- the operation of deleting the merged monitoring records may be performed after the parameters are read out, before the merge operation is completed, or after the merge operation is completed. The former method can release the memory space as soon as possible, the latter one. This approach can improve security, for example, to avoid data loss caused by power outages when the consolidated monitoring record has not been calculated.
- the two blocks of monitoring parameters can be used to determine the degree of heat and cold of the data block. If the IOPS and 10 counts in the two monitoring records are similar, the data blocks corresponding to the two monitoring records are considered to be similar in heat and cold. If they are all hot blocks, they can be migrated together to a faster read storage medium. If they are cold data, they can be migrated together to a slower read storage medium.
- the time point of cold and heat change, the time interval of cold and heat change, and the time interval may be The three monitoring parameters of the frequency of the access 10 to determine the hot and cold changes of the data block. If the difference between the three monitoring parameters is consistent with the respective preset thresholds, that is, the hot and cold change time points are similar, the hot and cold change time intervals are similar, and the bursty 10 times are similar. It is considered that the data blocks monitored by the two monitoring records have similar cold and hot variations, and the monitoring records of the data blocks can be combined and stored.
- Step 13 Write the combined monitoring record into the memory, and the combined monitoring record is used to monitor the data block monitored by the monitoring records before the merge.
- the monitoring record AB is generated by combining the monitoring record A and the monitoring record B in FIG. 2, and the broken line in FIG. 2 and FIG. 3 is the correspondence between the monitoring record and the data block. relationship.
- the number of in-memory monitoring records has been reduced from 5 in Figure 2 to 4 in Figure 3, reducing the memory footprint.
- Monitoring record The value of each monitoring parameter in AB and the monitoring record A are combined with the monitoring record in monitoring record B.
- the monitoring record AB can simultaneously monitor 10 requests for data block A and data block B.
- step S14 may be further included.
- Step S14 monitoring the host's I/O request for the data block in the memory, and combining the I/O request record of the data block monitored by the monitoring record for updating the monitoring parameters of the combined monitoring record in the memory, That is to say, each new 10 request, as long as it is a request for the data block monitored by the combined monitoring record, will be included in the monitoring parameters of the combined monitoring record.
- each new 10 request as long as it is a request for the data block monitored by the combined monitoring record, will be included in the monitoring parameters of the combined monitoring record.
- the merged monitoring record is written into the memory, and a new 10 request accesses the data block A or the access data block B, the new 10 request is counted into the monitoring record AB. go with.
- a split method of monitoring records may also be included.
- Split method one see step S15 of Fig. 4.
- the merged monitoring record can be re-opened and stored in the memory.
- the monitoring record generated by the disassembly is referred to as a split monitoring record. For example, when there is a lot of free space in the memory space, or the user wants to reduce Small monitoring granularity.
- the method of disassembling is: reading the merged monitoring record from the memory, splitting the combined monitoring record into at least one split monitoring record, and each of the split monitoring records is used for monitoring at least one data block,
- the value of each monitoring record in the split monitoring record is equal to the value of each monitoring record when the monitored data block is regarded as a data block.
- the combined monitoring record carries the identifier of the monitored data block.
- the splitting method is no longer unique.
- the combined monitoring record monitors 3 data blocks, and the combined monitoring record can be split into 3 monitoring records; each monitoring record monitors one data block; it can also be split into 2 monitoring records, one of which monitors 2 monitoring records. Data block, another monitoring record monitors 1 data block.
- the splitting of the monitoring record is the reverse process of combining the monitoring records.
- each monitoring parameter after splitting can be allocated according to the number of monitored data blocks.
- IOPS the monitoring parameter
- each monitoring parameter after splitting can be allocated according to the number of monitored data blocks.
- IOPS the monitoring parameter
- the splitting process may split the monitoring record of the combined monitoring record or the single data block without depending on the number of data blocks or the data block size monitored by the monitoring data.
- the disassembled generated monitoring record is called split monitoring record, which means that the method in Figure 5 can be executed together with Embodiments 1 and 2, or can be performed separately.
- the split monitoring record can carry the offset address and length of the monitored data, and can further carry the data block to which the monitoring data belongs.
- the sum of the monitoring parameter values is equal to the value of the monitoring parameter in the combined monitoring record.
- the value of each monitoring parameter may be allocated according to the percentage of the monitored data volume and the amount of data monitored by the split monitoring record, that is, carried in each monitoring record according to the splitting.
- the data length, the monitoring parameters are distributed to the respective monitoring records after the split.
- the split method shown in Figure 5 includes the following steps.
- Step 21 Read at least one monitoring parameter from each monitoring record in the memory.
- Step 22 Split each of the monitored monitoring parameters into a monitoring record that meets the preset threshold, and generate at least two monitoring records after the splitting, and each of the monitored monitoring records carries The offset address and the data length of the data.
- the sum of the monitoring parameter values is equal to the value of the monitoring parameter in the combined monitoring record.
- the ratio of the value of each monitoring parameter in the split monitoring record to the corresponding monitoring parameter in the split monitoring record may be the ratio of the monitored data of the split monitoring record to the amount of data monitored by the split monitoring record.
- Step 23 Write the post-split monitoring record into the memory, and the split monitoring record is used to monitor the data of the carried offset address and the data length.
- Each monitoring record also carries the offset address and length of the monitored data. The offset address is the starting address of the data. The monitored data can be found by offsetting the address and length.
- the randomness can be used as the basis for splitting the monitoring record: Analyze the monitoring records in the memory, find the two monitoring parameters of random 10 quantity and total 10 quantity, and take the ratio of random 10 quantity to total 10 quantity as randomness. For the data in the data block with high randomness, the access frequency of different parts may be quite different. For such monitoring records, in the case of sufficient memory, splitting can improve the monitoring accuracy.
- the random 10 quantity and the total 10 quantity are both monitoring parameters. Random 10 is relative to 10 consecutive times, referring to two different read/writes 10. The previous end address differs greatly from the previous start address.
- the user can specify a threshold, the previous end address and The difference between the next start address and the threshold is considered to be random 10; 10 consecutive reads of 10 different reads/writes 10, the previous end address is not much different from the previous start address.
- other parameters than the randomness may be used, for example, the monitoring record with the IOPS greater than the preset threshold is split, or the monitoring record with the 10 size and the 10 time delay are all in accordance with the respective thresholds. Minute.
- Embodiments of the present invention also provide a monitoring record management device, such as a memory controller or a program running in a memory controller, and the device can apply the above method.
- the storage system is composed of a storage controller, a memory, and a memory.
- the memory stores a plurality of monitoring records, each monitoring record monitors one data block in the memory, and one monitoring record is a set of monitoring parameters of a data block,
- the monitoring parameters include, for example, the number of read/write times per second of the monitored data block, the read/write frequency of the data block, and the read/write randomness of the data block.
- the monitor monitors the read and write requests of the data block and updates the monitoring records in the memory in real time with the monitoring results.
- the storage controller is physically independent of the storage.
- the storage server In the storage server, the storage server itself has storage control.
- the storage controller is integrated in the storage server, so the storage server itself is implemented. In other forms of storage systems, this method can be performed by a device having similar computing power as the storage controller. In the embodiment of the present invention, whether or not it is a physically independent storage controller, only having functions similar thereto are collectively referred to as a storage controller.
- the monitoring record management apparatus 5 includes a first reading module 51, a merging module 52 connected to the first reading module 51, and a first writing module 53 connected to the merging module 52, and the following for each functional module The function continues to be introduced one by one.
- the first reading module 51 is configured to read at least one monitoring parameter from each monitoring record of the memory.
- each monitoring record is used to monitor a data block in the memory, and each monitoring record is composed of at least one monitoring parameter, and each monitoring parameter is read and written by the monitoring request for 10 data blocks.
- the situation was obtained.
- there are a total of five monitoring records in the memory namely monitoring records, monitoring records, and monitoring records. , monitoring record D and monitoring record £.
- Each monitoring record monitors one data block and monitors a total of 5 data blocks.
- Each monitoring record includes three monitoring parameters.
- monitoring record A includes monitoring parameter A1, monitoring parameter A2, and monitoring parameter A3.
- the parameters can be, for example, I0PS, read/write frequency, and 10 counts, respectively.
- Whether the action of the first reading module is performed may be triggered by the user, or may be triggered by the storage system according to a preset threshold condition. For example, if the existing monitoring parameters are already full of memory, That is to say, when the memory has no extra space to store more monitoring records, the action of the first reading module is performed. At this time, if new data blocks need to be monitored, there will be cases where the monitoring parameters of the new data block are not available for use. Or the monitored data blocks are generally small, for example, only a few K. It is not necessary to separately set the monitoring parameters for each data block, and the action of the first reading module can also be performed.
- the merging module 52 is configured to read, from the memory, a monitoring record having one or more monitoring parameter differences that meet a preset threshold, and combine the read monitoring records to generate a combined monitoring record, where the combined monitoring record records The identifier of the monitored data block, the read monitoring record is deleted from the memory, wherein the value of each monitoring parameter in the combined monitoring record is equal to when the monitored multiple data blocks are regarded as one data block The value of each monitoring parameter.
- the read monitoring records are combined to generate a combined monitoring record, which is to merge the monitored monitoring records with one or more monitoring parameter differences that meet the preset threshold. .
- the threshold value of each monitoring parameter difference can be specified by the user.
- the merged monitoring record carries the identifier of the monitored data block, and the identifiers of different data blocks are different.
- each monitoring record Before the merger, each monitoring record also records the identification of the data block monitored by each.
- the combined monitoring record monitors a plurality of data blocks, so the identification of the plurality of data blocks is recorded in the combined monitoring record, and the identifiers of the different data blocks are different. Through the identification, the user can know which data is monitored by the monitoring parameter. Piece.
- the merging module 52 performs a threshold detection of the difference of the content read by the reading module 51.
- the number of monitored monitoring parameters may be one or more, that is, only one monitoring parameter may be judged. If the difference of the monitoring parameters meets the threshold in different monitoring records, a matching manner of the threshold is If it is less than or not greater than a specific value, the monitoring records are read out as a whole, and it is assumed that the monitoring parameters read in step 101 include IOPS, monitoring records 8, monitoring records ⁇ monitoring records C, monitoring records 0, monitoring records E
- the recorded IOPS are 50, 52, 60, 61, 70, respectively, and the IOPS has a threshold of 3, which means that the monitoring records with IOPS difference less than 3 are in compliance with the combined conditions.
- the monitoring parameters of the monitoring record A and the monitoring record B are read out and combined separately, and the combined monitoring records simultaneously record the monitoring records of the data block A and the data block B; the monitoring record C and the monitoring record D are also recorded.
- the memory will be read out and merged, and the combined monitoring record will record the monitoring records of data block C and data block D at the same time.
- the following examples are further based on the previous examples. It is assumed that in addition to the IOPS, the monitoring parameter 10 count (10 count) is compared, the monitoring parameters, Monitoring record ⁇ monitoring record, monitoring record 0, monitoring record E, 10 counts are 10, 11, 20, 30, 40, respectively, the threshold of 10 count is 5, that is, only monitoring parameter A and monitoring parameter B 10
- the monitoring record A and the monitoring record B satisfy the threshold requirements of IOPS and 10 counts at the same time.
- the monitoring record C and the monitoring record D satisfy the threshold requirement of IOPS, but do not satisfy the threshold requirement of 10 counts. Therefore, only monitor record A and monitor B will be read out and combined to generate a combined monitor record. Monitor record C and monitor record D will not be read from the memory.
- the value of the combined monitoring parameter is equal to the value of each monitoring parameter when the monitored data block is regarded as a data block, or the value of the monitoring parameter obtained when monitoring the monitored data block set, in the data block set.
- a single 10 request for any data block is recorded as a 10 request for the entire data block set.
- the value of the combined monitoring parameter is generated by combining the monitoring parameters of all monitored data blocks, and the data block monitored by the combined monitoring parameter is a collection of data blocks monitored by the combined monitoring parameters.
- the pre-merging monitoring records record the I/O request records of the respective monitored data blocks, and the consolidated monitoring records record the I/O request records of the monitored data block sets.
- the merged monitoring records are originally monitored with data blocks, and the values of the monitoring records in the combined monitoring parameters are equal to the values obtained by monitoring the data blocks together, and the monitoring records in the combined monitoring parameters may also be referred to as merges. Monitor parameters.
- the monitoring parameter A and the monitoring parameter B are all composed of three monitoring parameters, which are IOPS, 10 counts, and 10 frequencies, respectively, in which the IOPS of the monitoring record A is The value is 50, the value of the 10 count is 10, and the 10 frequency is 10; the value of the IOPS in the monitoring record B is 52, the value of the 10 count is 11, and the frequency of 10 is 20.
- the values of the monitoring parameters are directly added, because if the set consisting of the data block A and the data block B is regarded as a new data block AB, the IOPS is taken as an example, and the data block A or the data block is used.
- the operation of deleting the merged monitoring records may be performed after the parameters are read out, before the merge operation is completed, or after the merge operation is completed. The former method can release the memory space as soon as possible, the latter one. This approach can improve security, for example, to avoid data loss caused by power outages when the consolidated monitoring record has not been calculated.
- the two blocks of monitoring parameters can be used to determine the degree of heat and cold of the data block. If the IOPS and 10 counts in the two monitoring records are similar, the data blocks corresponding to the two monitoring records are considered to be similar in heat and cold. If they are all hot blocks, they can be migrated together to a faster read storage medium. If they are cold data, they can be migrated together to a slower read storage medium.
- the change rule of the data block can be judged by the three monitoring parameters of the cold and hot change time point, the cold and heat change time interval, and the bursting frequency 10 access frequency. If the difference between the three monitoring parameters meets the respective preset thresholds in the two monitoring records, that is, the hot and cold changing time points are similar, the hot and cold changing time intervals are similar, and the bursting 10 times are similar. It is considered that the data blocks monitored by the two monitoring records have similar cold and hot variations, and the monitoring records of these data blocks can be combined and stored.
- the first writing module 53 is configured to write the combined monitoring record into the memory, and the combined monitoring record is used to monitor the data block monitored by the monitoring records before the combining.
- the monitoring record AB is generated by combining the monitoring record A and the monitoring record B in FIG.
- the number of in-memory monitoring records has been reduced from 5 in Figure 2 to 4 in Figure 3, reducing the memory footprint.
- Monitoring record The value of each monitoring parameter in AB and the monitoring record A are combined with the monitoring record in monitoring record B.
- the monitoring record AB can simultaneously monitor 10 requests for data block A and data block B.
- the writing module can also be used to monitor the host's I/O request for the data block in the memory, and use the I/O request record of the data block monitored by the combined monitoring record. Update the monitoring parameters of the combined monitoring record in the memory. That is to say, each new 10 request, as long as it is a request for the data block monitored by the combined monitoring record, will be included in the monitoring parameters of the combined monitoring record.
- the specific example in the fourth embodiment as an example, if the merged monitoring record is written into the memory and a new 10 request accesses the data block A or the access data block B, the new 10 request is counted into the monitoring record AB. go with.
- the splitting module 54 connected to the first write module 53 may be further included. See Figure 6. After the monitoring record is merged, the splitting module 54 can re-open the merged monitoring record and store it in the memory if necessary. In this embodiment, the split monitoring record is called a split monitoring record. For example, when there is a lot of free space in the memory space, or when the user wants to reduce the monitoring granularity.
- the function of the splitting module 54 is to read the merged monitoring record from the memory, split the combined monitoring record into at least one split monitoring record, and each of the split monitoring records is used to monitor at least one data. Block, the value of each monitoring record in the split monitoring record is equal to the value of each monitoring record when the monitored data block is regarded as one data block.
- the combined monitoring record carries the identifier of the monitored data block.
- the splitting method is no longer unique.
- the combined monitoring record monitors 3 data blocks, and the combined monitoring record can be split into 3 monitoring records; each monitoring record monitors one data block; it can also be split into 2 monitoring records, one of which monitors 2 monitoring records. Data block, another monitoring record monitors 1 data block.
- the splitting of the monitoring record is the reverse process of combining the monitoring records.
- the splitting method can be various. For example, after the splitting, each monitoring parameter can follow the monitored data block. The amount is apportioned.
- the splitting process may split the monitoring record of the combined monitoring record or the single data block without depending on the number of data blocks or the data block size monitored by the monitoring data. Disassembling the generated monitoring records is called split monitoring records.
- the split monitoring record can carry the offset address and length of the monitored data, and can further carry the data block to which the monitoring data belongs.
- the sum of the monitoring parameter values is equal to the value of the monitoring parameter in the combined monitoring record.
- the value of each monitoring parameter may be allocated according to the percentage of the monitored data volume and the amount of data monitored by the split monitoring record, that is, carried in each monitoring record according to the splitting. The data length, the monitoring parameters are distributed to the respective monitoring records after the split.
- the data block splitting apparatus shown in Fig. 7 includes a second reading module 61, a second splitting module 62 connected to the second reading module, and a second writing module 63 connected to the second splitting module.
- the second reading module 61 is configured to read at least one monitoring parameter from each monitoring record of the memory, where the monitoring record may be a combined monitoring record or a combined monitoring record.
- the second splitting module 62 is configured to split the monitored monitoring parameters, each monitoring parameter meets the preset threshold value, and generate at least two monitoring records after splitting, after each splitting
- the monitoring record carries the offset address and the data length of the data.
- the sum of the monitoring parameter values is equal to the value of the monitoring parameter in the combined monitoring record.
- the ratio of the value of each monitoring parameter in the split monitoring record to the corresponding monitoring parameter in the split monitoring record may be the ratio of the monitored data of the split monitoring record to the amount of data monitored by the split monitoring record.
- the second writing module writes the post-split monitoring record into the memory, and the split monitoring record is used for Monitor the data of the offset address and data length carried.
- Each monitoring record also carries the offset address and length of the monitored data. The offset address is the starting address of the data. The monitored data can be found by offsetting the address and length.
- the randomness can be used as the basis for splitting the monitoring record: Analyze the monitoring records in the memory, find the two monitoring parameters of random 10 quantity and total 10 quantity, and take the ratio of random 10 quantity to total 10 quantity as randomness. For the data in the data block with high randomness, the access frequency of different parts may be quite different. For such monitoring records, in the case of sufficient memory, splitting can improve the monitoring accuracy.
- the random 10 quantity and the total 10 quantity are both monitoring parameters. Random 10 is relative to 10 consecutive times, referring to two different read/writes 10. The previous end address differs greatly from the previous start address.
- the user can specify a threshold, the previous end address and The difference between the next start address and the threshold is considered to be random 10; 10 consecutive reads of 10 different reads/writes 10, the previous end address is not much different from the previous start address.
- other parameters than the randomness may be used, for example, the monitoring record with the IOPS greater than the preset threshold is split, or the monitoring record with the 10 size and the 10 time delay are all in accordance with the respective thresholds. Minute.
- each monitoring parameter meets a preset threshold, which may be that the monitoring parameter itself meets a threshold value, or the monitoring parameter may be converted to a threshold value after being calculated by a preset algorithm, for example, the difference between the two monitoring parameters is met.
- the threshold, or the ratio of the two monitored parameters corresponds to the threshold.
- the combined monitoring parameters monitor only one data block.
- the merged monitoring records can be merged again, and those skilled in the art will understand that since the re-merging is only a repeated application of the foregoing embodiment of the present invention, it will not be described in detail.
- the monitoring after splitting can be split again.
- the technical solution of the present invention which is essential or contributes to the prior art, may be embodied in the form of a software product stored in a readable storage medium, such as a floppy disk of a computer.
- a hard disk or optical disk or the like includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform the methods described in various embodiments of the present invention.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Hardware Design (AREA)
- Mathematical Physics (AREA)
- Quality & Reliability (AREA)
- Computing Systems (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Debugging And Monitoring (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
Description
Claims
Priority Applications (5)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP12877608.5A EP2811410B1 (en) | 2012-12-21 | 2012-12-21 | Monitoring record management method and device |
CN201280006844.6A CN103502925B (zh) | 2012-12-21 | 2012-12-21 | 一种监控记录管理方法与装置 |
JP2015548142A JP6060276B2 (ja) | 2012-12-21 | 2012-12-21 | 監視レコード管理方法及びデバイス |
PCT/CN2012/087170 WO2014094303A1 (zh) | 2012-12-21 | 2012-12-21 | 一种监控记录管理方法与装置 |
US14/140,151 US8924642B2 (en) | 2012-12-21 | 2013-12-24 | Monitoring record management method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/CN2012/087170 WO2014094303A1 (zh) | 2012-12-21 | 2012-12-21 | 一种监控记录管理方法与装置 |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US14/140,151 Continuation US8924642B2 (en) | 2012-12-21 | 2013-12-24 | Monitoring record management method and device |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2014094303A1 true WO2014094303A1 (zh) | 2014-06-26 |
Family
ID=49866740
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2012/087170 WO2014094303A1 (zh) | 2012-12-21 | 2012-12-21 | 一种监控记录管理方法与装置 |
Country Status (5)
Country | Link |
---|---|
US (1) | US8924642B2 (zh) |
EP (1) | EP2811410B1 (zh) |
JP (1) | JP6060276B2 (zh) |
CN (1) | CN103502925B (zh) |
WO (1) | WO2014094303A1 (zh) |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9607036B2 (en) * | 2013-08-21 | 2017-03-28 | International Business Machines Corporation | Managing a data set |
CN105574055B (zh) * | 2014-11-07 | 2019-02-26 | 阿里巴巴集团控股有限公司 | 防止内存耗尽的方法及装置 |
CN105095099B (zh) * | 2015-07-21 | 2017-12-29 | 浙江大学 | 一种基于内存页位图变更的大内存页整合方法 |
CN105022696B (zh) * | 2015-07-21 | 2017-12-29 | 浙江大学 | 一种基于内存访问热度的大内存页整合方法 |
US10659532B2 (en) * | 2015-09-26 | 2020-05-19 | Intel Corporation | Technologies for reducing latency variation of stored data object requests |
CN117331511B (zh) * | 2023-11-30 | 2024-02-23 | 苏州元脑智能科技有限公司 | 一种存储设备及其数据传输方法、装置、系统和存储介质 |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020129048A1 (en) * | 2000-03-03 | 2002-09-12 | Surgient Networks, Inc. | Systems and methods for resource monitoring in information storage environments |
US20110099268A1 (en) * | 2009-10-26 | 2011-04-28 | Hitachi, Ltd. | Information processing system, and management method for storage monitoring server |
US20110144951A1 (en) * | 2009-12-11 | 2011-06-16 | Lieberman Donald A | Monitoring memory module parameters in high performance computers |
US20110197046A1 (en) * | 2010-02-05 | 2011-08-11 | International Business Machines Corporation | Storage application performance matching |
US20120131196A1 (en) * | 2010-11-18 | 2012-05-24 | Hitachi, Ltd. | Computer system management apparatus and management method |
Family Cites Families (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP4252102B2 (ja) * | 1993-06-21 | 2009-04-08 | 株式会社日立製作所 | 計算機システムおよび二次記憶装置 |
US6125392A (en) * | 1996-10-11 | 2000-09-26 | Intel Corporation | Method and apparatus for high speed event log data compression within a non-volatile storage area |
US20020049841A1 (en) * | 2000-03-03 | 2002-04-25 | Johnson Scott C | Systems and methods for providing differentiated service in information management environments |
US9268780B2 (en) * | 2004-07-01 | 2016-02-23 | Emc Corporation | Content-driven information lifecycle management |
JP4579000B2 (ja) * | 2005-02-14 | 2010-11-10 | 株式会社日立製作所 | 計算機システムにおけるデータ配置設定 |
US20080250057A1 (en) * | 2005-09-27 | 2008-10-09 | Rothstein Russell I | Data Table Management System and Methods Useful Therefor |
JP4749140B2 (ja) * | 2005-12-05 | 2011-08-17 | 株式会社日立製作所 | データマイグレーション方法及びシステム |
JP4284332B2 (ja) * | 2006-04-21 | 2009-06-24 | 株式会社東芝 | パフォーマンスモニタ装置、データ収集方法及びそのプログラム |
WO2008007348A1 (en) * | 2006-07-12 | 2008-01-17 | Mpstor Limited | A data storage system |
US7496707B2 (en) * | 2006-08-22 | 2009-02-24 | International Business Machines Corporation | Dynamically scalable queues for performance driven PCI express memory traffic |
US8627017B2 (en) * | 2008-12-30 | 2014-01-07 | Intel Corporation | Read and write monitoring attributes in transactional memory (TM) systems |
JP4832553B2 (ja) * | 2009-06-16 | 2011-12-07 | 株式会社日立製作所 | コピー開始タイミングを制御するシステム及び方法 |
US8489844B2 (en) * | 2009-12-24 | 2013-07-16 | Hitachi, Ltd. | Storage system providing heterogeneous virtual volumes and storage area re-allocation |
US8578092B2 (en) * | 2010-01-25 | 2013-11-05 | Hitachi, Ltd. | Method and apparatus to support determining storage area unit size |
US20140122774A1 (en) * | 2012-10-31 | 2014-05-01 | Hong Kong Applied Science and Technology Research Institute Company Limited | Method for Managing Data of Solid State Storage with Data Attributes |
-
2012
- 2012-12-21 JP JP2015548142A patent/JP6060276B2/ja active Active
- 2012-12-21 WO PCT/CN2012/087170 patent/WO2014094303A1/zh active Application Filing
- 2012-12-21 EP EP12877608.5A patent/EP2811410B1/en active Active
- 2012-12-21 CN CN201280006844.6A patent/CN103502925B/zh active Active
-
2013
- 2013-12-24 US US14/140,151 patent/US8924642B2/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020129048A1 (en) * | 2000-03-03 | 2002-09-12 | Surgient Networks, Inc. | Systems and methods for resource monitoring in information storage environments |
US20110099268A1 (en) * | 2009-10-26 | 2011-04-28 | Hitachi, Ltd. | Information processing system, and management method for storage monitoring server |
US20110144951A1 (en) * | 2009-12-11 | 2011-06-16 | Lieberman Donald A | Monitoring memory module parameters in high performance computers |
US20110197046A1 (en) * | 2010-02-05 | 2011-08-11 | International Business Machines Corporation | Storage application performance matching |
US20120131196A1 (en) * | 2010-11-18 | 2012-05-24 | Hitachi, Ltd. | Computer system management apparatus and management method |
Non-Patent Citations (1)
Title |
---|
See also references of EP2811410A4 * |
Also Published As
Publication number | Publication date |
---|---|
CN103502925A (zh) | 2014-01-08 |
EP2811410B1 (en) | 2018-05-30 |
JP2016500459A (ja) | 2016-01-12 |
EP2811410A1 (en) | 2014-12-10 |
US20140181332A1 (en) | 2014-06-26 |
CN103502925B (zh) | 2016-05-25 |
US8924642B2 (en) | 2014-12-30 |
EP2811410A4 (en) | 2017-04-05 |
JP6060276B2 (ja) | 2017-01-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10346081B2 (en) | Handling data block migration to efficiently utilize higher performance tiers in a multi-tier storage environment | |
WO2021073635A1 (zh) | 一种数据存储方法及装置 | |
US10250684B2 (en) | Methods and systems for determining performance capacity of a resource of a networked storage environment | |
JP5090941B2 (ja) | ストレージサブシステム及びストレージシステム | |
WO2014094303A1 (zh) | 一种监控记录管理方法与装置 | |
WO2013178155A1 (zh) | 分布式文件系统的存储方法及装置 | |
CN103116661B (zh) | 一种数据库的数据处理方法 | |
US10509739B1 (en) | Optimized read IO for mix read/write scenario by chunking write IOs | |
CA2977742C (en) | Method for deduplication in storage system, storage system, and controller | |
CN104516824B (zh) | 数据存储系统中的存储管理方法和系统 | |
WO2017206754A1 (zh) | 分布式文件系统的存储方法和存储装置 | |
CN108089825B (zh) | 一种基于分布式集群的存储系统 | |
US20110185140A1 (en) | Method and apparatus to support determining storage area unit size | |
EP2981920B1 (en) | Detection of user behavior using time series modeling | |
WO2016041127A1 (zh) | 重复数据删除方法和存储阵列 | |
JP6582721B2 (ja) | 制御装置、ストレージシステム、及び制御プログラム | |
WO2019120226A1 (zh) | 数据访问预测方法和装置 | |
US11249666B2 (en) | Storage control apparatus | |
WO2023050856A1 (zh) | 数据处理方法及存储系统 | |
WO2017122263A1 (ja) | 管理計算機及び管理方法 | |
CN107357532A (zh) | 一种新型集群存储新型缓存预读实现方法 | |
WO2023155703A1 (zh) | 一种工作负载特征提取方法及装置 | |
US20230305728A1 (en) | Management system, storage system, and management processing method | |
JP2021131752A (ja) | 情報処理装置及び情報処理プログラム | |
CN109992211A (zh) | 一种文件数据管理方法及相关装置 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 2012877608 Country of ref document: EP |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 12877608 Country of ref document: EP Kind code of ref document: A1 |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 12877608 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2015548142 Country of ref document: JP Kind code of ref document: A |
|
NENP | Non-entry into the national phase |
Ref country code: DE |