CN103488544B - Method and apparatus for detecting slow processing tray - Google Patents

Method and apparatus for detecting slow processing tray Download PDF

Info

Publication number
CN103488544B
CN103488544B CN 201310445133 CN201310445133A CN103488544B CN 103488544 B CN103488544 B CN 103488544B CN 201310445133 CN201310445133 CN 201310445133 CN 201310445133 A CN201310445133 A CN 201310445133A CN 103488544 B CN103488544 B CN 103488544B
Authority
CN
Grant status
Grant
Patent type
Prior art keywords
time interval
time
request
slow
corresponding
Prior art date
Application number
CN 201310445133
Other languages
Chinese (zh)
Other versions
CN103488544A (en )
Inventor
叶丞
李元
毛长生
Original Assignee
华为技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Grant date

Links

Abstract

本发明实施例提供一种检测慢盘的处理方法和装置,该包括:从预置检测时间段的起始时刻开始获取硬盘上各个输入/输出I/O请求的I/O延迟;根据各个I/O请求的I/O延迟,确定各个I/O请求与各个时间区间的对应关系;各个时间区间是将预置检测时间段的时间长度划分为多个区间得到的;在预置检测时间段的结束时刻,根据各个I/O请求与各个时间区间的对应关系,确定各个时间区间对应的I/O请求数量;根据各个时间区间对应的I/O请求数量和各个时间区间对应的预设阈值,确定硬盘为慢盘。 Example embodiments provide a processing method and apparatus for detecting slow disk of the present invention, which includes: the hard disk starts acquiring the respective input / output I / O request I / O time delay from the start of the pre-detection period; I according to the respective / O request I / O delay, determine respective I / O request correspondence relationship between the respective time interval; each time interval a preset length of time is divided into a plurality of detection period sections obtained; preset detection period end time, each I / O request correspondence relationship between the respective time interval in accordance with determining respective time interval corresponding to the I / O request quantity; number of requests, and each time interval corresponding to a preset threshold according to the respective time interval corresponding to the I / O to determine the hard disk is slow. 通过该检测慢盘的处理方法和装置,充分考虑了I/O请求数量与I/O延迟的关联关系,从而可准确的检测出慢盘并移除该慢盘,避免慢盘对业务的处理造成不良影响。 By means of the detection and treatment of slow disk full consideration of I / O requests association with the number I / O delay, thereby accurately detect and remove the slow slow disk drive, the disk to avoid slow processing of the traffic adverse effects.

Description

检测慢盘的处理方法和装置 Method and apparatus for detecting slow processing tray

技术领域 FIELD

[0001]本发明实施例涉及计算机技术,尤其涉及一种检测慢盘的处理方法和装置。 Example embodiments relate to computer technology [0001] The present invention particularly relates to a processing method and apparatus for detecting slow disc.

背景技术 Background technique

[0002] 为了满足不断提高业务的处理效率,设备上通常设置多个硬盘,但硬盘由于坏道和外界振动等原因,处理1/0(Input/Output,输入/输出)请求时会花较多的时间进行纠错,导致对I/O请求响应过慢,从而该硬盘上所处理的业务的处理效率降低,严重时造成主机业务中断,影响了设备的性能,其中这种对I/O请求响应过慢,影响业务处理效率的硬盘称为慢盘。 [0002] In order to meet the increasing business efficiency, usually a plurality of hard disk devices, but because of bad sectors and external hard disk vibration, treatment 1/0 (Input / Output, Input / Output) request will spend more time error correction, resulting in an I / O request response is too slow, so that the processing efficiency of operations on the hard disk of the processing is reduced, resulting in severe host service interruption, affect the performance of the device, wherein such requests for I / O response is too slow, affecting business efficiency of the hard disk called slow. 因此,在设备通过多个硬盘处理业务时,实时的对各个硬盘进行检测,具体的,记录各个硬盘上每个I/O请求的I/O延迟,所述I/O延迟为I/O请求从下发给硬盘的起始时刻至该硬盘对为I/O请求进行响应所需的时间;随后统计在一段时间内I/O延迟超过预设时间门限值的I/O请求数量,若一个硬盘上I/O延迟超过预设时间门限值的I/O请求数量超过一定的预设阈值,则确定该硬盘为慢盘,则便将该慢盘从磁盘阵列的RAID(Redundant Arrays ofInexpensive Disks,磁盘阵列)组中移除,以保障设备的处理业务的效率。 Accordingly, when the business processing apparatus through a plurality of hard disks, hard real-time detection of the respective, specific recorded on each hard disk each I / O request I / O delay, the I / O delay of I / O requests starting time sent from the hard disk to the hard disk in response to the time required for the I / O request; statistics over time then I / O delay exceeds a preset time threshold I / O request number, if a hard disk I / O time delay exceeds a preset threshold of I / O requests quantity exceeds a certain preset threshold, it is determined that the slower hard disk, then the disk slowly from the disk array of RAID (Redundant arrays ofInexpensive disks, RAID) group is removed, to ensure the efficiency of business processing apparatus.

[0003] 但在实际应用中发现,硬盘上I/O延迟与I/O请求数量的关联关系,即在一段时间内,硬盘上I/O延迟较低的I/O请求的数量较多,而I/O延迟较高的I/O请求的数量较少,因此,若一个硬盘上I/O延迟小于300ms(毫秒)的I/O请求数量为60,而I/O延迟处于300ms〜400ms之间的I/O请求数量为2,若此时判断慢盘的判断条件为“I/O延迟超过300ms的I/O请求数量超过50时,该硬盘为慢盘”,则依据该判断条件,会确认上述硬盘不是慢盘;但设备在处理业务时,硬盘上I/O延迟较高的I/O请求的数量尽管很少,但对硬盘处理业务的效率所产生的不良影响却较大,因此,采用现有技术中的检测方案,不能准确的检测出慢盘,不能及时的将慢盘移除,导致设备的性能下降。 [0003] However, in practical applications found on the disk I / O delays and I / O relationship of the number of requests, i.e. a period of time, the disk I / O low latency high I / O requests, and I / O latency high low I / O requests, and therefore, if a hard disk I / O latency is less than 300ms (milliseconds) of the I / O request quantity is 60, and the I / O delay is 300ms~400ms between the I / O request quantity is 2, if the determination at this time is slower disk determination condition "I / O delays exceed 300ms of I / O requests when the number exceeds 50, the hard disk is slow," it is determined according to the condition , will confirm the above slower than hard disk; however, in dealing with traffic device, the disk I / O delays adverse effect of higher I / O requests although the number is small, but the efficiency of the processing operations on the hard drive has generated a large Therefore, the detection scheme employed in the prior art can not accurately detect slow disk, the disk can not be removed in time will be slower, resulting in performance degradation of the device.

发明内容 SUMMARY

[0004]本发明实施例提供一种检测慢盘的处理方法和装置,用于优化慢盘检测的过程,以使设备上硬盘的布置合理。 [0004] Example embodiments provide a processing method and apparatus for detecting slow disc for disc detection slow process optimization of the present invention, so that the device is arranged on the hard disk reasonable.

[0005]第一方面,本发明实施例提供一种检测慢盘的处理方法,包括: [0005] In a first aspect, embodiments of the present invention provides a method for detecting slow disk processing, comprising:

[0006]从预置检测时间段的起始时刻开始获取硬盘上各个输入/输出I/O请求的I/O延迟; [0006] hard disk starts acquiring the respective input / output I / O request I / O time delay from the start of the pre-detection period;

[0007]根据各个所述I/O请求的I/O延迟,确定各个所述I/O请求与各个时间区间的对应关系;所述各个时间区间是将所述预置检测时间段的时间长度划分为多个区间得到的; [0007] According to each of the I / O request I / O delay, determining the correspondences between the I / O request to the respective time interval; each time the time length of the interval is a preset detection time period divided into a plurality of sections obtained;

[0008]在所述预置检测时间段的结束时刻,根据各个所述I/O请求与各个所述时间区间的对应关系,确定各个所述时间区间对应的I/o请求数量; [0008] In the preset end time of the detection period, according to each of the I / O request correspondence between each of said time intervals, determining a time interval corresponding to each of the I / o request quantity;

[0009]根据各个所述时间区间对应的I/O请求数量和各个所述时间区间对应的预设阈值,确定所述硬盘为慢盘。 [0009] The time interval corresponding to each of the I / O requests and the number of each of the time intervals corresponding to a preset threshold value, determines that the hard disk is slow.

[0010]结合第一方面,在第一实施方式中,所述根据各个所述I/O请求的I/O延迟,确定各个所述I/O请求与所述预置检测时间段内的各个时间区间的对应关系,包括: [0010] with the first aspect, in the first embodiment, according to the each of the I / O request I / O delay, determining each of the respective I / O request to the preset detection time period time interval corresponding relationship, comprising:

[0011 ]比较所述I/O请求的I/O延迟与各个所述时间区间的起点和终点; [0011] Comparison of the I / O request I / O delay of the start and end of each time interval;

[0012]若所述I/O请求的I/O延迟大于所述各个所述时间区间内第一时间区间的起点,并小于所述第一时间区间的终点,则确定所述I/o请求对应于所述第一时间区间。 [0012] When the I / O request I / O delay is greater than each of the starting point of the time interval within a first time interval, and the end is less than a first time interval, determining that the I / o requests corresponding to said first time interval.

[0013]结合第一方面第一实施方式,在第二实施方式中,所述根据各个所述时间区间对应的I/o请求数量和各个所述时间区间对应的预设阈值,确定所述硬盘为慢盘,包括: [0013] connection with the first aspect of the first embodiment, in the second embodiment, each of the number of requests and the time interval corresponding to a predetermined threshold according to the time interval corresponding to the respective I / o, the hard drive is determined slow disk, including:

[0014]若所述时间区间对应的I/O请求数量大于或等于所述时间区间对应的预设阈值,则确定所述时间区间为超限区间; [0014] If the time interval corresponding to the I / O request quantity greater than or equal to the time interval corresponding to a preset threshold value, it is determined that the time interval is overrun interval;

[0015]在所述各个时间区间中,若存在超限区间,则确定所述硬盘为慢盘;或者 [0015] In the respective time interval, if the overrun interval exists, it is determined that the hard disk is slow; or

[0016]所述各个时间区间中的每一个时间区间都是所述超限区间,则确定所述硬盘为慢盘; [0016] Each of the respective time interval is the time interval overrun interval, determining that the hard disk is slow;

[0017]结合第一方面至第一方面第二实施方式中任意一种实施方式,在第三实施方式中,所述预置检测时间段的起始时刻与向所述硬盘下发各个I/o请求的下发时刻相同。 [0017] with the first aspect to the first aspect of the second embodiment, any one embodiment, in the third embodiment, the pre-detection period start time at the hard disk and sent to each I / o the same time issued the request.

[0018]第二方面,本发明实施例提供一种检测慢盘的处理装置,包括: [0018] a second aspect, embodiments of the present invention provides apparatus for detecting slow disk processing, comprising:

[0019]获取模块,用于从预置检测时间段的起始时刻开始获取硬盘上各个输入/输出I/O请求的I/o延迟; [0019] The acquisition module for acquiring the respective input on the hard disk starts / output I / O request I / o from the start time of the preset delay detection period;

[0020]第一确定模块,用于根据各个所述I/O请求的I/O延迟,确定各个所述I/O请求与各个时间区间的对应关系;所述各个时间区间是将所述预置检测时间段的时间长度划分为多个区间得到的; [0020] a first determining module for each of the I / O request I / O delay, determine each of the I / O relationship between the respective time interval corresponds to a request; the respective time interval is the pre- dividing the time length of the detection period set for the plurality of sections obtained;

[0021]第二确定模块,用于在所述预置检测时间段的结束时刻,根据各个所述I/O请求与各个所述时间区间的对应关系,确定各个所述时间区间对应的I/o请求数量; [0021] The second determining module, configured to detect the preset end time period, according to each of the I / O request correspondence between each of said time interval, the time interval determining each of the corresponding I / o request quantity;

[0022]第三确定模块,用于根据各个所述时间区间对应的I/O请求数量和各个所述时间区间对应的预设阈值,确定所述硬盘为慢盘。 [0022] The third determining module, according to the time interval corresponding to the respective I / O requests and the number of each of the time intervals corresponding to a preset threshold value, determines that the hard disk is slow.

[0023]结合第二方面,在第一实施方式中,第一确定模块具体用于比较所述I/O请求的I/O延迟与各个所述时间区间的起点和终点; [0023] combination with the second aspect, in the first embodiment, the first determining module is configured to compare the I / O request I / O delay of the start and end of each time interval;

[0024]若所述I/O请求的I/O延迟大于所述各个所述时间区间内第一时间区间的起点,并小于所述第一时间区间的终点,则确定所述I/O请求对应于所述第一时间区间。 [0024] When the I / O request I / O delay is greater than each of the starting point of the time interval within a first time interval, and the end is less than a first time interval, determining that the I / O request corresponding to said first time interval.

[0025]结合第二方面第一实施方式,在第二实施方式中,第三确定模块具体用于若所述时间区间对应的I/O请求数量大于或等于所述时间区间对应的预设阈值,则确定所述时间区间为超限区间; [0025] conjunction with the second aspect of the first embodiment, in the second embodiment, the third module is configured to determine if the time interval corresponding to the I / O request quantity greater than or equal to the time interval corresponding to a preset threshold value , it is determined that the time interval is overrun interval;

[0026]在所述各个时间区间中,若存在超限区间,则确定所述硬盘为慢盘;或者 [0026] In the respective time interval, if the overrun interval exists, it is determined that the hard disk is slow; or

[0027]所述各个时间区间中的每一个时间区间都是所述超限区间,则确定所述硬盘为慢盘; [0027] Each of the respective time interval is the time interval overrun interval, determining that the hard disk is slow;

[0028]结合第二方面至第二方面第二实施方式中任意一种实施方式,在第三实施方式中,所述预置检测时间段的起始时刻与向所述硬盘下发各个I/o请求的下发时刻相同。 [0028] bound to a second aspect of the second embodiment of the second aspect of any one embodiment, in the third embodiment, the pre-detection period start time and sent to each of the said hard disk I / o the same time issued the request.

[0029]本发明实施例提供的检测慢盘的处理方法和装置,通过将预置检测时间段的时间长度划分为多个时间区间,并向每一个时间区间设置了预设阈值,从而可以根据各个时间区间对应的I/O请求数量和各个时间区间对应的预设阈值,准确的确定慢盘,充分考虑了I/O请求数量与I/O延迟的关联关系,即在时间范畴上考虑了I/O请求的分布规律,从而准确的检测出慢盘并移除该慢盘,避免慢盘对业务的处理造成不良影响,以提高设备的性能。 [0029] The processing method and apparatus for detecting slow disk according to an embodiment of the present invention, by dividing the length of the preset time period is detected a plurality of time intervals, and the interval is set each time a preset threshold value, may be calculated according each time interval corresponding to the I / O request quantity and the respective time intervals corresponding to a preset threshold value, accurate determination of slow disk full consideration of I / O requests association the number of I / O delay, i.e. the time category considered distribution of I / O requests, thereby accurately detect and remove the slow slow disk drive, to avoid adversely affecting the business process slow disk, to improve the performance of the device.

附图说明 BRIEF DESCRIPTION

[0030]为了更清楚地说明本发明实施例或现有技术中的技术方案,下面将对实施例或现有技术描述中所需要使用的附图作一简单地介绍,显而易见地,下面描述中的附图是本发明的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动性的前提下,还可以根据这些附图获得其他的附图。 [0030] In order to more clearly illustrate the technical solutions in the embodiments or the prior art embodiment of the present invention, the accompanying drawings for illustrating the prior art described or needed to be used in an embodiment will be briefly introduced hereinafter, the description below the figures show some embodiments of the present invention, those of ordinary skill in the art is concerned, without any creative effort, and can obtain other drawings based on these drawings.

[0031 ]图1为本发明检测慢盘的处理方法实施例一的流程图; [0031] FIG. 1 is a flowchart of detecting slow disk processing method embodiment of the invention;

[0032]图2为本发明检测慢盘的处理方法实施例二的流程图; [0032] FIG 2 the processing method for detecting slow disc according to a second embodiment of a flow diagram of the present invention;

[0033]图3为本发明检测慢盘的处理装置实施例一的结构图; [0033] FIG. 3 slower processing means detects a disc configuration diagram of the embodiment of the present invention;

[0034]图4为本发明检测慢盘的处理装置实施例二的结构图。 [0034] FIG. 4 processing means for detecting slow disk configuration diagram according to a second embodiment of the present invention.

具体实施方式 detailed description

[0035]为使本发明实施例的目的、技术方案和优点更加清楚,下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例是本发明一部分实施例,而不是全部的实施例。 [0035] In order that the invention object, technical solutions, and advantages of the embodiments more clearly, the following the present invention in the accompanying drawings, technical solutions of embodiments of the present invention are clearly and completely described, obviously, the described the embodiment is an embodiment of the present invention is a part, but not all embodiments. 基于本发明中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都属于本发明保护的范围。 Based on the embodiments of the present invention, all other embodiments of ordinary skill in the art without any creative effort shall fall within the scope of the present invention.

[0036]图1为本发明检测慢盘的处理方法实施例一的流程图。 [0036] FIG. 1 processing method for detecting a disc slower flow diagram embodiment of the present invention. 如图1所示,本实施例的执行主体为检测慢盘的处理装置,该装置可以采用硬件和/或软件的方式实现,优选的,该处理装置可以设置在装载了硬盘的设备上,便于对设备中的硬盘进行检测,具体的,该方法包括: As shown, the embodiment according to the present embodiment performs the main processing for the slower disk detecting apparatus, the apparatus can be implemented in hardware and / or software implemented 1, preferably, the processing means may be provided in a device loaded on the hard disk, to facilitate a hard disk of the detecting device, specifically, the method comprising:

[0037] S101、从预置检测时间段的起始时刻开始获取硬盘上各个I/O请求的I/O延迟。 [0037] S101, the hard disk starts acquiring each I / O request I from the preset starting time detection period / O delay.

[0038]通常来说,设备通过驱动程序、I/O接口等向设备中的硬盘下发I/O请求,且设备在向硬盘下发的各个I/o请求时,会对各个I/O请求打上时间戳,该时间戳记录了I/O请求下发的时刻,随着时间的推移,当硬盘对I/O请求做出响应时,上述处理装置会根据硬盘对I/O请求做出响应的时刻和该I / O请求上的时间戳记录的下发时刻,计算出各个I / O请求的I / O延迟。 [0038] Generally speaking, by the device drivers, I / O interface to the device delivers a hard disk I / O request, and when the device / o requests issued by each of the hard disk I, will each I / O request time stamped, the time stamp issued by the recording I / O requests, over time, when responding to the hard disk I / O request, the processing device will make a hard disk according to the I / O request transmitting time stamp recorded on the response time and the I / O request is calculated each I / O request I / O latency.

[0039]其中所预置检测时间段的起始时刻与向硬盘下发各个I/O请求的下发时刻相同,即当前需要对硬盘进行检测,则设备向硬盘下发各个I/o请求的同时,上述处理装置启动预置时间段计时,该启动时刻即为预置检测时间段的起始时刻。 [0039] wherein the preset starting time detection period and transmitting time issued to the hard disk each I / O request is the same as that currently required for detecting hard disk, the device delivers the respective hard disk I / o requests Meanwhile, the processing device start counting a preset time, the activation time is the preset start time of the detection period. 其中预置时间段的具体时间长度可根据实际需要灵活设置,通常根据实际经验,以设备上出现过的最高I/o延迟的时间长度作为预置时间段的时间长度。 Wherein the specific length of time preset time period may be set flexibly according to actual needs, generally based on actual experience, the length of time the device appeared to maximum I / o a preset length of time as the delay period.

[0040] S102、根据各个I/O请求的I/O延迟,确定各个I/O请求与各个时间区间的对应关系O [0040] S102, in accordance with various I / O request I / O delay, determine respective I / O requests to the respective time interval corresponding relationship O

[0041]上述各个时间区间是将所述预置检测时间段的时间长度划分为多个区间得到的;具体的,预置时间段的是时间长度为T,则将该时间长度T划分为[XhX2)、[χ2,χ3)……[X1,Xi+1)……[Xn-l,Xn),其中,Xn=T,i为正整数,且n>i M,且Xi〈Xi+lo [0041] The respective time intervals are preset time length of the detection period is divided into a plurality of sections obtained; Specifically, the preset time period is a time length T, the duration T is divided into the [ XhX2), [χ2, χ3) ...... [X1, Xi + 1) ...... [Xn-l, Xn), where, Xn = T, i is a positive integer, and n> i M, and Xi <Xi + lo

[0042]当处理装置获取一个I/O请求的I/O延迟,则比较该I/O请求的I/O延迟与各个所述时间区间的起点和终点;若I/O请求的I/O延迟大于各个时间区间内第一时间区间的起点(上述X1),并小于第一时间区间的终点(χ1+1),则确定所述I/O请求对应于所述第一时间区间,该第一时间区间可以为上述各个时间区间内任意一个时间区间,对于一个I/o请求来说,若存在一个时间区间的起点小于该I/o请求的I/O延迟,且该时间区间的终点大于该I/O请求的I/o延迟,则该时间区间对于该I/O请求来说为第一时间区间。 [0042] When the processing apparatus acquires an I / O request I / O delay, then I / O delay the start and end of each time interval of the I / O request comparison; if I / O request I / O delay is greater than the start of the first time interval in each time interval (the x1 above), and the end is less than the first time interval (χ1 + 1), it is determined that the I / O request corresponding to said first time interval, the second a time interval may be any time interval within said respective time interval, for an I / o requests, if there is a time interval smaller than the start of the I / o requested I / O delay, and the end of the time interval is greater than the I / O request I / o delay, the time interval for the I / O request for a first time interval.

[0043] 举例来说,预置时间段T=400ms,该预置时间段被划分为[O,100ms)、[100ms,200ms)、[200ms,300ms)、[300ms ,400ms);当一个I/O请求的I/O延迟为250ms时,通过比较该I/O请求的I/O延迟与各个所述时间区间的起点和终点,确定第一时间区间[200ms,300ms)的起点200ms小于250ms,且终点300ms大于250ms,则确定该I/O请求对应于第一时间区间[200ms,300ms)。 [0043] For example, the preset time period T = 400ms, the preset time period is divided into [O, 100ms), [100ms, 200ms), [200ms, 300ms), [300ms, 400ms); When an I when / O request I / O delay is 250ms, by comparing the I / O request I / O latency to each of the start and end time interval, determining a first time interval [200ms, 300ms) the starting point is less than 200ms 250ms , and the end point is greater than 300ms 250ms, it is determined that the I / O request corresponding to a first time interval [200ms, 300ms).

[0044] S103、在预置检测时间段的结束时刻,根据各个I/O请求与各个所述时间区间的对应关系,确定各个时间区间对应的I/o请求数量; [0044] S103, a preset end time of the detection period, in accordance with various I / O request correspondence between each of said time intervals, each time interval corresponding to the determined I / o request quantity;

[0045]如SlOl中所述,处理装置在设备下发I/O请求时开始进行计时,且各个I/O请求被响应的时刻也不同,则随着时间的推移,每当一个I/O请求被响应,则处理装置执行SlOl及S102,确定出各个I/O请求与各个时间区间的对应关系,则当时钟到达预置检测时间段的结束时刻,根据各个I/o请求与各个时间区间的对应关系,确定各个时间区间对应的I/O请求数量,例如,在预置检测时间段内,依次有I/o请求I至I/O请求5被响应,则通过SlOl和S102,确定I/o请求1、I/O请求2和I/O请求5共3个I/O请求对应于时间区间[200ms,300ms),而I/O请求3和I/O请求4对应于[300ms,400ms),则通过S103,确定时间区间[200ms,300ms)对应的I/O请求个数为3,而时间区间[300ms,400ms)对应的I/O请求个数为2。 [0045] As described SlOl, the processing apparatus made in device I / O request to start measuring time, and each I / O request response time is also different, over time, each time an I / O request is acknowledged, the process execution means SlOl and S102, it is determined that each of the I / O request correspondence relationship between the respective time interval, the time when the clock reaches the end of the pre-detection period, in accordance with various I / o requests each time interval correspondence between each determined time interval corresponding to the I / O request number, e.g., in the preset detection time period, there are sequentially I / o I to request I / O request is in response to 5, and SlOl through S102, determines the I / o requests 1, I / O request 2 and I / O request 5 a total of three I / O request corresponding to the time interval [200ms, 300ms), and the I / O request 3, and I / O request 4 corresponding to [300ms, 400ms), through S103, determining the time interval [200ms, 300ms) corresponding to the I / O requests to the number 3, and the time interval [300ms, 400ms) corresponding to the I / O requests is the number 2.

[0046] S104、根据各个时间区间对应的I/O请求数量和各个时间区间对应的预设阈值,确定所述硬盘为慢盘。 [0046] S104, in accordance with various time intervals corresponding to I / O requests and the number of time intervals each corresponding to a preset threshold value, determines that the hard disk is slow.

[0047]本实施例中,处理装置对向每个时间区间设置了预设阈值,则在确定硬盘是否为慢盘时,是根据各个所述时间区间对应的I/O请求数量和各个所述时间区间对应的预设阈值,确定硬盘是否为慢盘,充分考虑了I/O请求数量与I/O延迟的关联关系,即在时间范畴上考虑了I/O请求的分布规律,从而可以准确的确定硬盘是否为慢盘。 [0047] In this embodiment, the processing section of the apparatus is provided each time to a preset threshold, it is determined whether the hard disk when the disk is slow, and a number of requests according to each time interval corresponding to each of the I / O of the time interval corresponding to a preset threshold value, it is determined whether the hard disk is slower disk full consideration to the I / O request association the number of I / O delay, i.e. the time visible to consider the distribution of the I / O request, which can accurately to determine whether the hard disk is slow.

[0048]本实施例中,通过将预置检测时间段的时间长度划分为多个时间区间,并向每一个时间区间设置了预设阈值,从而可以根据各个时间区间对应的I/O请求数量和各个时间区间对应的预设阈值,准确的确定慢盘,充分考虑了I/O请求数量与I/O延迟的关联关系,SP在时间范畴上考虑了I/O请求的分布规律,从而可以准确的检测出慢盘并移除该慢盘,避免慢盘对业务的处理造成不良影响,以提高设备的性能。 [0048] In this embodiment, the length of time by a preset detection period is divided into a plurality of time intervals, and the interval is set each time a preset threshold, which can request quantity according to a time interval corresponding to the respective I / O and a preset threshold value for each time interval corresponding to the accurate determination of slow disk, full account of the relationship I / O requests the number of I / O delays, SP at the time visible to consider the distribution of the I / O request, which can accurate detection of slow disk and remove the disk slow, slow pan to avoid adverse effects on the business process to improve the performance of the device.

[0049]图2为本发明检测慢盘的处理方法实施例二的流程图。 [0049] FIG 2 the processing method for detecting slow disc flowchart of a second embodiment of the present invention. 如图2所示,该方法是在图1所示的实施例一的基础上,进一步描述检测慢盘的处理方法,具体的,该方法包括: 2, which is a basis of the embodiment shown in FIG. 1, the processing method is further described disc detecting slow, particularly, the method comprising:

[0050] S201、从预置检测时间段的起始时刻开始获取硬盘上各个输入/输出I/O请求的I/O延迟。 [0050] S201, the hard disk starts acquiring the respective input / output I from the start time preset detection period / O request I / O latency.

[0051 ] 具体如SlOl所述,在此不再赘述。 [0051] DETAILED SlOl as said, are not repeated here.

[0052 ] S202、比较I /0请求的I /0延迟与各个时间区间的起点和终点。 [0052] S202, Comparative I / I / 0 to the respective delay time interval 0 start and end requests.

[0053]即由S201中所述的预置检测时间段的起始时刻起,每当一个I/O请求被响应后便执行S201,随后根据该I/O请求的I/O延迟,与各个时间区间的起点和终点进行比较。 [0053] i.e., from the starting time of the pre-detection period from the step S201, every time an I / O request is executed after S201 in response, then the delay based on the I / O request I / O, and the respective time interval start and end points are compared.

[0054] S203、若I/O请求的I/O延迟大于各个所述时间区间内第一时间区间的起点,并小于第一时间区间的终点,则确定I/o请求对应于所述第一时间区间。 [0054] S203, if the I / O request I / O delay is greater than a starting point in each time interval of said first time interval and less than the end of the first time interval, it is determined that I / o corresponding to the first request time interval.

[0055]具体的,假设I/O请求5被响应,根据I/O请求5的I/O延迟,依次与预置时间段内的各个时间区间的起点和终点比较,例如I/O请求5的I/O延迟大于一个时间区间[X1An)的起点Xi并小于该时间区间[Xi,Xi+i)的终点Xi+i,即该时间区间[Xi,Xi+i)对I/O请求5来说为第一时间区间,则确定该I/O请求对应于该第一时间区间,具体的,可以将该时间区间[X1,χ1+1)对应的计数器加I;当然可以理解的,若I/o请求5的I/O延迟不满足条件“不大于一个时间区间[Xi,Xi+i)的起点Xi并小于该时间区间[Xi,Xi+i)的终点Xi+i”,则该时间区间[Xi,Xi+i)对应的计数器保持不变,则I/O请求5的I/O延迟继续与其他时间区间去比较;可选的,在确定出I/o请求与某个时间区间对应时,可以在时间区间表中,将该I/O请求记录在对应的时间区间所在的表格内。 [0055] Specifically, assume that I / O request is in response to 5, I 5 according to the requested I / O / O delay, and sequentially comparing the start and end of each time interval a preset period of time, for example, I / O requests 5 the I / O delay is longer than a time interval [X1An) starting point Xi and smaller than the time interval [Xi, Xi + i) of end point Xi + i, i.e., the time interval [Xi, Xi + i) of the I / O request 5 for a first time interval, it is determined that the I / O request corresponding to the first time interval. specifically, the time interval can be [x1, χ1 + 1) corresponding to the counter by I; of course be appreciated that, if I / o I 5 request / O delay condition is not satisfied "is not larger than a time interval [Xi, Xi + i) is smaller than the starting point of the time interval and Xi [Xi, Xi + i) of end point Xi + i", the time interval [Xi, Xi + i) corresponding to the counter remains unchanged, the I / O request 5 of delayed I / O continues to compare with other time intervals; optionally, in determining the I / o requests a certain time when the corresponding interval, the interval time table may be, the I / O request corresponding to the record in the table where the time interval.

[0056] S204、在预置检测时间段的结束时刻,根据各个I/O请求与各个时间区间的对应关系,确定各个时间区间对应的I/o请求数量。 [0056] S204, a preset end time of the detection period, in accordance with various I / O request corresponding relation to various time intervals, each time interval corresponding to the determined I / o request quantity.

[0057]基于S203中的描述,各个I/O请求与各个时间区间的对应关系可以通过各个时间区间的计数器,或时间区间表中各个时间区间所在的表格内记录的I/o请求来表征,则在预置检测时间段的结束时刻,处理装置在执行S204时,可以查看各个时间区间的计数器,或者统计时间区间表中各个时间区间所在的表格内记录的I/O请求的数量,以确定各个时间区间对应的I /0请求数量。 [0057] Based on in S203 described, each I / O can be I / o requests through the respective time interval counter, or time interval table each time recorded in the table section where the characterizing request correspondence relationship between the respective time interval, at the end of the preset time of the detection period, the processing unit when executing S204, the counter can view each time interval, time interval, or statistical table where each time interval table record I / O request number, to determine time interval corresponding to respective I / 0 request quantity.

[0058] S205、根据各个时间区间对应的I/O请求数量和各个所述时间区间对应的预设阈值,确定所述硬盘为慢盘。 [0058] S205, in accordance with various time intervals corresponding to I / O requests and the number of each of the time intervals corresponding to a preset threshold value, determines that the hard disk is slow.

[0059]具体的,比较时间区间对应的I/O请求数量和各个时间区间对应的预设阈值,若所述各个时间区间内的一个时间区间对应的I/o请求数量大于或等于该时间区间对应的预设阈值,则确定该时间区间为超限区间;随后在确定硬盘是否为慢盘时,可以依据的第一种判断条件为: [0059] Specifically, the comparison time interval corresponding to the I / O requests and the number of time intervals corresponding to the respective predetermined threshold value, if the time interval corresponding to a respective time interval within the I / o is greater than or equal to the requested number of time interval corresponding to a preset threshold value, it is determined that the time interval is overrun interval; then determining whether the hard disk is slower, may be based on a first determination condition is:

[0060]在各个时间区间中,若存在超限区间,则确定硬盘为慢盘; [0060] In each time interval, if the overrun interval exists, it is determined that the hard disk is slow;

[0061 ]在实际应用中,慢盘对某个设备的影响较大,则在针对该设备检测慢盘时,可以以较严格的条件检测,即各个时间区间中,一旦存在超限区间,便确定该硬盘为慢盘。 [0061] In practice, the greater impact on a slow disk device, then when the device for detecting slow disk, you can detect a more stringent conditions, that is, each time interval, once there is overrun range, they determined that the hard disk is slow. 以上述划分的时间区间举例,时间区间[O,100ms)对应的I/O请求数量为60且预设阈值为100,时间区间[100ms,200ms)对应的I/O请求数量为80且预设阈值为80、[ 200ms,300ms)对应的I/O请求数量为30且预设阈值为50、[300ms,400ms)对应的I/O请求数量为2且预设阈值为2,其中超限区间为[10ms,200ms)和[300ms,400ms),即在四个时间区间中存在了两个超限区间,则确定该硬盘为慢盘,其中各个时间区间的预设阈值是考虑了出现延迟的I/o请求数量的分布规律,例如I/o延迟较高的I/O请求数量较少,但I/O延迟较高的I/O请求数量较少,因此对应的,表征较高延迟的时间区间[300mS,400mS)设置的预设阈值则较小,以便可准确的检测出慢盘; A time interval of the partition example, the time interval [O, 100ms) corresponding to the I / O request quantity of 60 100 and a preset threshold value, the time interval [100ms, 200ms) corresponding to the I / O requests and the preset quantity of 80 threshold 80, [200ms, 300ms) corresponding to the I / O requests to the number 30 and the preset threshold is 50, [300ms, 400ms) corresponding to the I / O requests is the number 2 and the preset threshold value 2, wherein the gauge section is [10ms, 200ms) and [300ms, 400ms), i.e. in the presence of the four time interval overrun two sections, it is determined that the slower hard disk, wherein each time interval a preset threshold is considered the delays I / o number of distribution requests, for example, I / o high latency I / O requests fewer, but the I / O latency high I / O requests a smaller number, thus the corresponding, characterized by higher latency time interval [300mS, 400mS) set a preset threshold value is small, so that it can accurately detect slow disks;

[0062]灵活的,还可以采用采用第二种判断条件: [0062] The flexible, may also be employed using the second judgment condition:

[0063]各个时间区间中的每一个时间区间都是超限区间,则确定硬盘为慢盘。 [0063] Each time interval in each time interval is overrun range, it is determined that the hard disk is slow. 在实际中,有些硬盘从实质上来说,不是慢盘,但在检测过程中,有时设备会因为其他原因出现死机等极端情况,以使某个时间区间内的I/O请求数量超过该时间区间的预设阈值,导致出现一个超限区间,但其他时间区间都不是超限区间,则处理装置不会将该硬盘误判为慢盘。 In practice, some of the hard disk in essence, not slow the disc, but in the detection process, the device sometimes extreme crash situation like for other reasons, so that within a certain time interval I / O request quantity exceeds the time interval the preset threshold value, resulting in a gauge section, but other time intervals are not overrun range, the processing device will not be mistaken for the slower hard disk.

[0064]需要说明的是,由于设备中有多个硬盘,所述处理装置可针对每个硬盘开启检测进程,即针对每个硬盘去执行上述实施例中的各个步骤,以在各个硬盘中确定出慢盘。 [0064] Incidentally, since a plurality of hard disk devices, said processing means detecting process for each of the openable hard disk, i.e., to perform the above-described embodiments of the various steps for each hard disk, the hard disk to determine each a slow disk.

[0065]本实施例中,通过将预置检测时间段的时间长度划分为多个时间区间,并向每一个时间区间设置了预设阈值,从而可以根据各个时间区间对应的I/O请求数量和各个时间区间对应的预设阈值,准确的确定慢盘,充分考虑了I/O请求数量与I/O延迟的关联关系,SP在时间范畴上考虑了I/O请求的分布规律,从而准确的检测出慢盘并移除该慢盘,避免慢盘对业务的处理造成不良影响,,以提高设备的性能。 [0065] In this embodiment, the preset time length of the detection period is divided into a plurality of time intervals, and the interval is set each time a preset threshold, which can request quantity according to a time interval corresponding to the respective I / O and each time interval corresponding to a preset threshold value, accurate determination of slow disk full consideration of I / O requests association the number of I / O delays, SP at the time visible to consider the distribution of the I / O request to accurately the detection of a slow disk and remove the disk slow, slow pan to avoid adverse effects on the business process ,, to improve the performance of the device.

[0066]图3为本发明检测慢盘的处理装置实施例一的结构图。 [0066] FIG. 3 slower processing means detects a disc configuration diagram of the embodiment of the present invention. 如图3所示,该装置包括: As shown in FIG 3, the apparatus comprising:

[0067]获取模块31,用于从预置检测时间段的起始时刻开始获取硬盘上各个输入/输出I/O请求的I/O延迟; [0067] The obtaining module 31, a hard disk starts acquiring the respective input / output I / O request I / O time delay from the start of the pre-detection period;

[0068]第一确定模块32,用于根据各个所述I/O请求的I/O延迟,确定各个所述I/O请求与各个时间区间的对应关系;所述各个时间区间是将所述预置检测时间段划分为多个区间得到的; [0068] The first determining module 32, according to each of the I / O request I / O delay, determine each of the I / O request correspondence relationship between the respective time interval; the respective time interval is the preset detection time period into a plurality of sections obtained;

[0069]第二确定模块33,用于在所述预置检测时间段的结束时刻,根据各个所述I/O请求与各个所述时间区间的对应关系,确定各个所述时间区间对应的I/o请求数量; [0069] The second determination module 33 in the preset end time for the detection period, according to each of the I / O request correspondence between each of said time intervals, determining a time interval corresponding to each of the I / o request quantity;

[0070]第三确定模块34,用于根据各个所述时间区间对应的I/O请求数量和各个所述时间区间对应的预设阈值,确定所述硬盘为慢盘。 [0070] The third determining module 34, according to the time interval corresponding to the respective I / O requests and the number of each of the time intervals corresponding to a preset threshold value, determines that the hard disk is slow.

[0071]进一步的,上述第一确定模块32具体用于比较所述I/O请求的I/O延迟与各个所述时间区间的起点和终点; [0071] Further, the first determining module 32 is configured to compare the I / O request I / O latency to each of the start and end time interval;

[0072]若所述I/O请求的I/O延迟大于所述各个所述时间区间内第一时间区间的起点,并小于所述第一时间区间的终点,则确定所述I/O请求对应于所述第一时间区间。 [0072] When the I / O request I / O delay is greater than each of the starting point of the time interval within a first time interval, and the end is less than a first time interval, determining that the I / O request corresponding to said first time interval.

[0073]进一步的,第三确定模块34具体用于若所述时间区间对应的I/O请求数量大于或等于所述时间区间对应的预设阈值,则确定所述时间区间为超限区间; [0073] Further, the third determining module 34 is specifically configured to, if a time interval corresponding to the I / O request quantity greater than or equal to the time interval corresponding to a preset threshold value, it is determined that the time interval is overrun interval;

[0074]在所述各个时间区间中,若存在超限区间,则确定所述硬盘为慢盘;或者 [0074] In the respective time interval, if the overrun interval exists, it is determined that the hard disk is slow; or

[0075]所述各个时间区间中的每一个时间区间都是所述超限区间,则确定所述硬盘为慢盘; [0075] Each of the respective time interval is the time interval overrun interval, determining that the hard disk is slow;

[0076]其中,上述预置检测时间段的起始时刻与向所述硬盘下发各个I/O请求的下发时亥IJ相同。 [0076] wherein the above-described starting time preset detection time period is the same as the Hai IJ issued to the hard disk of each I / O request when the hair.

[0077]需要说明的是,上述各个模块对应执行前述各个方法实施例中的各个步骤,在此不再赘述。 [0077] Incidentally, the above-described respective modules corresponding to the various execution method in the embodiment the various steps, are not repeated herein.

[0078]本发明实施例提供的检测慢盘的处理装置,通过将预置检测时间段的时间长度划分为多个时间区间,并向每一个时间区间设置了预设阈值,从而可以根据各个时间区间对应的I/O请求数量和各个时间区间对应的预设阈值,准确的确定慢盘,充分考虑了I/O请求数量与I/O延迟的关联关系,即在时间范畴上考虑了I/O请求的分布规律,从而准确的检测出慢盘并移除该慢盘,避免慢盘对业务的处理造成不良影响,以提高设备的性能。 [0078] processing means for detecting slow disk according to an embodiment of the present invention, the length of time by a preset detection period is divided into a plurality of time intervals, and the interval is set each time a preset threshold, according to the respective time can be interval corresponding I / O request quantity and the respective time intervals corresponding to a preset threshold value, accurate determination of slow disk full consideration of I / O requests association the number of I / O delay, i.e. the time visible considered I / distribution of O request, thereby accurately detect and remove the slow slow disk drive, to avoid adversely affecting the business process slow disk, to improve the performance of the device.

[0079]图4为本发明检测慢盘的处理装置实施例二的结构图。 [0079] FIG. 4 processing means for detecting slow disk configuration diagram according to a second embodiment of the present invention. 如图4所示,该装置具体用于执行前述各个方法实施例中的各个步骤,具体该装置包括: As shown in FIG 4, the apparatus particularly for performing the aforementioned method in the various steps of the various embodiments, particularly the apparatus comprises:

[0080]存储器41,用于存储数据信息,该数据信息包括从预置检测时间段的起始时刻开始获取硬盘上各个输入/输出I/o请求的I/O延迟; [0080] The memory 41, for storing data information, the information includes a start acquiring data on the hard disk of each input / output I / o I request from the preset starting time detection period / O delay;

[0081]处理器42,用于根据各个所述I/O请求的I/O延迟,确定各个所述I/O请求与各个时间区间的对应关系;所述各个时间区间是将所述预置检测时间段的时间长度划分为多个区间得到的; [0081] The processor 42, according to each of the I / O request I / O delay, determine each of the I / O request corresponding to the respective time interval relationship; each said preset time interval is the duration detection period is divided into a plurality of sections obtained;

[0082]处理器42,用于在所述预置检测时间段的结束时刻,根据各个所述I/O请求与各个所述时间区间的对应关系,确定各个所述时间区间对应的I/o请求数量; [0082] processor 42 for detecting the preset end time period, according to each of the I / O request correspondence between each of said time intervals, determining a time interval corresponding to each of the I / o request quantity;

[0083]处理器42,用于根据各个所述时间区间对应的I/O请求数量和各个所述时间区间对应的预设阈值,确定所述硬盘为慢盘。 [0083] The processor 42, according to the time interval corresponding to the respective I / O requests and the number of each of the time intervals corresponding to a preset threshold value, determines that the hard disk is slow.

[0084] 进一步的,上述处理器42具体用于比较所述I/O请求的I/O延迟与各个所述时间区间的起点和终点; [0084] Further, the above-described processor 42 for comparing the particular I / O request I / O delay of the start and end of each time interval;

[0085]若所述I/O请求的I/O延迟大于所述各个所述时间区间内第一时间区间的起点,并小于所述第一时间区间的终点,则确定所述I/O请求对应于所述第一时间区间。 [0085] When the I / O request I / O delay is greater than each of the starting point of the time interval within a first time interval, and the end is less than a first time interval, determining that the I / O request corresponding to said first time interval.

[0086]进一步的,处理器42具体用于若所述时间区间对应的I/O请求数量大于或等于所述时间区间对应的预设阈值,则确定所述时间区间为超限区间; [0086] Further, if the processor 42 for the particular time interval corresponding to the I / O request quantity greater than or equal to the time interval corresponding to a preset threshold value, it is determined that the time interval is overrun interval;

[0087]在所述各个时间区间中,若存在超限区间,则确定所述硬盘为慢盘;或者 [0087] In the respective time interval, if the overrun interval exists, it is determined that the hard disk is slow; or

[0088]所述各个时间区间中的每一个时间区间都是所述超限区间,则确定所述硬盘为慢盘; [0088] Each of the respective time interval is the time interval overrun interval, determining that the hard disk is slow;

[0089]详细的,所述预置检测时间段的起始时刻与向所述硬盘下发各个I/O请求的下发时刻相同。 [0089] Specifically, the pre-detection period start time is the same to the hard disk each issued I / O request transmitting time.

[0090]本发明实施例提供的检测慢盘的处理装置,通过将预置检测时间段的时间长度划分为多个时间区间,并向每一个时间区间设置了预设阈值,从而可以根据各个时间区间对应的I/O请求数量和各个时间区间对应的预设阈值,准确的确定慢盘,充分考虑了I/O请求数量与I/o延迟的关联关系,即在时间范畴上考虑了I/O请求的分布规律,从而准确的检测出慢盘并移除该慢盘,避免慢盘对业务的处理造成不良影响,以提高设备的性能。 [0090] processing means for detecting slow disk according to an embodiment of the present invention, the length of time by a preset detection period is divided into a plurality of time intervals, and the interval is set each time a preset threshold, according to the respective time can be interval corresponding I / O request quantity and the respective time intervals corresponding to a preset threshold value, accurate determination of slow disk, full account of the relationship I / O requests the number of I / o delay, i.e. the time visible considered I / distribution of O request, thereby accurately detect and remove the slow slow disk drive, to avoid adversely affecting the business process slow disk, to improve the performance of the device.

[0091]本领域普通技术人员可以理解:实现上述各方法实施例的全部或部分步骤可以通过程序指令相关的硬件来完成。 [0091] Those of ordinary skill in the art will be appreciated that: each of the foregoing methods to complete all or part of the hardware can step through the program instructions relating to the implementation. 前述的程序可以存储于一计算机可读取存储介质中。 The program may be stored in a computer readable storage medium. 该程序在执行时,执行包括上述各方法实施例的步骤;而前述的存储介质包括:R0M、RAM、磁碟或者光盘等各种可以存储程序代码的介质。 When the program is executed, comprising the step of performing the above-described method of the embodiment; and the storage medium comprising: a variety of medium may store program codes R0M, RAM, magnetic disk, or optical disk.

[0092]最后应说明的是:以上各实施例仅用以说明本发明的技术方案,而非对其限制;尽管参照前述各实施例对本发明进行了详细的说明,本领域的普通技术人员应当理解:其依然可以对前述各实施例所记载的技术方案进行修改,或者对其中部分或者全部技术特征进行等同替换;而这些修改或者替换,并不使相应技术方案的本质脱离本发明各实施例技术方案的范围。 [0092] Finally, it should be noted that: the above embodiments only describe the technical solutions in embodiments of the present invention, rather than limiting;. Although the embodiments of the present invention has been described in detail, those of ordinary skill in the art should appreciated: it still may be made to the technical solutions described embodiments modifications, or to some or all of the technical features equivalents; as such modifications or replacements do not cause the essence of corresponding technical solutions to depart from embodiments of the present invention range of technical solutions.

Claims (6)

  1. 1.一种检测慢盘的处理方法,其特征在于,包括: 从预置检测时间段的起始时刻开始获取硬盘上各个输入/输出I/o请求的I/O延迟;根据各个所述I/o请求的I/O延迟,确定各个所述I/O请求与各个时间区间的对应关系;所述各个时间区间是将所述预置检测时间段的时间长度划分为多个区间得到的; 在所述预置检测时间段的结束时刻,根据各个所述I/O请求与各个所述时间区间的对应关系,确定各个所述时间区间对应的I/o请求数量; 根据各个所述时间区间对应的I/o请求数量和各个所述时间区间对应的预设阈值,确定所述硬盘为慢盘; 所述根据各个所述I/o请求的I/O延迟,确定各个所述I/O请求与所述预置检测时间段内的各个时间区间的对应关系,包括: 比较所述I/o请求的I/O延迟与各个所述时间区间的起点和终点; 若所述I/o请求的I/O延迟大于所述各个所述时间区间 A method for detecting slow processing tray, wherein, comprising: a hard disk starts acquiring the respective input / output I / o I request from the preset starting time detection period / O delay; I according to the respective / o requested I / O delay, determine each of the I / O request correspondence relationship between the respective time interval; the respective time interval is preset time length of the detection period is divided into a plurality of sections obtained; in the preset end time of the detection period, according to each of the I / O request correspondence between each of said time intervals, determining a time interval corresponding to each of the I / o request quantity; said time interval according to various corresponding to the I / o requests and the number of each of the time intervals corresponding to a preset threshold value, determines that the hard disk is slow; the delay according to each of the I / o requested I / O, each of the determined I / O request to the preset correspondence relationship between the respective time interval detection time period, comprising: comparing the I / o requested I / O delay the start and end of each time interval; and if the I / o requests the I / O delay is greater than the time period of each of the 第一时间区间的起点,并小于所述第一时间区间的终点,则确定所述I/o请求对应于所述第一时间区间。 Starting a first time interval, and the end is less than a first time interval, determining that the I / o request corresponding to said first time interval.
  2. 2.根据权利要求1所述的处理方法,其特征在于,所述根据各个所述时间区间对应的I/O请求数量和各个所述时间区间对应的预设阈值,确定所述硬盘为慢盘,包括: 若所述时间区间对应的I/o请求数量大于或等于所述时间区间对应的预设阈值,则确定所述时间区间为超限区间; 在所述各个时间区间中,若存在超限区间,则确定所述硬盘为慢盘;或者所述各个时间区间中的每一个时间区间都是所述超限区间,则确定所述硬盘为慢盘。 2. The processing method according to claim 1, wherein each of said time interval according to the corresponding I / O requests and the number of each of the time intervals corresponding to a preset threshold value, determines that the hard disk is slow , comprising: if the time interval corresponding to the I / o is greater than or equal to the requested number of time intervals corresponding to a preset threshold value, it is determined that the time interval is overrun interval; at the respective time interval, if present, over limited range, it is determined that the hard disk is slow; each of the respective time interval or time interval is the gauge section, determine that the hard disk is slow.
  3. 3.根据权利要求1或2所述的处理方法,其特征在于,所述预置检测时间段的起始时刻与向所述硬盘下发各个I/o请求的下发时刻相同。 The processing method of claim 1 or claim 2, wherein the preset detection period start time is the same to the hard disk each issued I / o request transmitting time.
  4. 4.一种检测慢盘的处理装置,其特征在于,包括: 获取模块,用于从预置检测时间段的起始时刻开始获取硬盘上各个输入/输出I/o请求的I/O延迟; 第一确定模块,用于根据各个所述I/o请求的I/O延迟,确定各个所述I/O请求与各个时间区间的对应关系;所述各个时间区间是将所述预置检测时间段的时间长度划分为多个区间得到的; 第二确定模块,用于在所述预置检测时间段的结束时刻,根据各个所述I/O请求与各个所述时间区间的对应关系,确定各个所述时间区间对应的I/o请求数量; 第三确定模块,用于根据各个所述时间区间对应的I/o请求数量和各个所述时间区间对应的预设阈值,确定所述硬盘为慢盘; 第一确定模块具体用于比较所述I/o请求的I/O延迟与各个所述时间区间的起点和终占.V , 若所述I/o请求的I/O延迟大于所述各个所述时间区间内第一时间区间 4. A method for detecting slow disc processing apparatus comprising: acquiring means for a preset time starting from the detection period of the hard disk starts acquiring the respective input / output I / o requested I / O delay; a first determining module for each of the I / o requested I / O delay, determine each of the I / O request correspondence relationship between the respective time interval; the respective time interval to the preset detection time is divided by the length of time period into a plurality of sections obtained; second determination means for detecting the end time of the preset period of time, according to each of the I / O request correspondence between each of said time interval, is determined time interval corresponding to each of the I / o request quantity; third determination module for each time interval corresponding to the I / o requests and the number of each of the time intervals corresponding to a preset threshold value, determines the hard disk slow pan; the I first determining module for comparing the I / o requests / O delay to each of the starting point and the final time interval accounted .V, if the I / o requested I / O delay is greater than the each of said first time interval within the time interval 起点,并小于所述第一时间区间的终点,则确定所述I/o请求对应于所述第一时间区间。 The starting point and the end point is smaller than a first time interval, determining that the I / o request corresponding to said first time interval.
  5. 5.根据权利要求4所述的处理装置,其特征在于,第三确定模块具体用于若所述时间区间对应的I/O请求数量大于或等于所述时间区间对应的预设阈值,则确定所述时间区间为超限区间; 在所述各个时间区间中,若存在超限区间,则确定所述硬盘为慢盘;或者所述各个时间区间中的每一个时间区间都是所述超限区间,则确定所述硬盘为慢盘。 The processing apparatus according to claim 4, characterized in that the third module is configured to determine if the time interval corresponding to the I / O request quantity greater than or equal to the time interval corresponding to a preset threshold value, it is determined the time interval is overrun interval; at the respective time interval, if the overrun interval exists, it is determined that the hard disk is slow; the or each respective time interval is the time interval overrun interval, it is determined that the hard disk is slow.
  6. 6.根据权利要求4或5所述的处理装置,其特征在于,所述预置检测时间段的起始时刻与向所述硬盘下发各个I/O请求的下发时刻相同。 The processing apparatus of claim 4 or claim 5, wherein the preset detection period start time is the same to the hard disk each issued I / O request transmitting time.
CN 201310445133 2013-09-26 2013-09-26 Method and apparatus for detecting slow processing tray CN103488544B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN 201310445133 CN103488544B (en) 2013-09-26 2013-09-26 Method and apparatus for detecting slow processing tray

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN 201310445133 CN103488544B (en) 2013-09-26 2013-09-26 Method and apparatus for detecting slow processing tray

Publications (2)

Publication Number Publication Date
CN103488544A true CN103488544A (en) 2014-01-01
CN103488544B true CN103488544B (en) 2016-08-17

Family

ID=49828799

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 201310445133 CN103488544B (en) 2013-09-26 2013-09-26 Method and apparatus for detecting slow processing tray

Country Status (1)

Country Link
CN (1) CN103488544B (en)

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103810062B (en) * 2014-03-05 2015-12-30 华为技术有限公司 Method and apparatus for detecting slow disk
WO2016029481A1 (en) * 2014-08-30 2016-03-03 华为技术有限公司 Method and device for isolating disk regions
CN106354590A (en) * 2015-07-17 2017-01-25 中兴通讯股份有限公司 Method and apparatus for disk detection
CN106407051A (en) * 2015-07-31 2017-02-15 华为技术有限公司 Slow disk detection method and device
CN106407052A (en) * 2015-07-31 2017-02-15 华为技术有限公司 Disk detection method and apparatus
CN106557389A (en) * 2015-09-29 2017-04-05 成都华为技术有限公司 Slow disk detection method and apparatus
CN105573888A (en) * 2015-12-14 2016-05-11 曙光信息产业股份有限公司 Method and apparatus for detecting disk performance in distributed file system
CN105630416A (en) * 2015-12-24 2016-06-01 创新科存储技术(深圳)有限公司 Disc kicking method and device in cloud storage system

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1955940A (en) * 2005-10-28 2007-05-02 富士通株式会社 RAID system, RAID controller and rebuilt/copy back processing method thereof
CN101625629A (en) * 2008-07-08 2010-01-13 日立环球储存科技荷兰有限公司 Techniques for scheduling requests for accessing storage devices using sliding windows
CN101770346A (en) * 2010-01-04 2010-07-07 浪潮电子信息产业股份有限公司 Method for dispatching multi-controller disk array coordination

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6779063B2 (en) * 2001-04-09 2004-08-17 Hitachi, Ltd. Direct access storage system having plural interfaces which permit receipt of block and file I/O requests

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1955940A (en) * 2005-10-28 2007-05-02 富士通株式会社 RAID system, RAID controller and rebuilt/copy back processing method thereof
CN101625629A (en) * 2008-07-08 2010-01-13 日立环球储存科技荷兰有限公司 Techniques for scheduling requests for accessing storage devices using sliding windows
CN101770346A (en) * 2010-01-04 2010-07-07 浪潮电子信息产业股份有限公司 Method for dispatching multi-controller disk array coordination

Also Published As

Publication number Publication date Type
CN103488544A (en) 2014-01-01 application

Similar Documents

Publication Publication Date Title
US20090006745A1 (en) Accessing snapshot data image of a data mirroring volume
US7434012B1 (en) Techniques for media scrubbing
US20100115211A1 (en) Behavioral monitoring of storage access patterns
CN102156738A (en) Method for processing data blocks, and data block storage equipment and system
US20100274933A1 (en) Method and apparatus for reducing memory size and bandwidth
US20120151148A1 (en) Systems and methods for background destaging storage tracks
US7882323B2 (en) Scheduling of background scrub commands to reduce high workload memory request latency
US20140075265A1 (en) Outputting Information of ECC Corrected Bits
CN102147708A (en) Method and device for detecting discs
US6735672B2 (en) Data storage array device and data access method
US20120272030A1 (en) Synchronous extent migration protocol for paired storage
US20130054913A1 (en) Storage system, and apparatus and method for controlling storage
US20110153928A1 (en) Memory utilization tracking
CN101840364A (en) Method for recovering data and storage device thereof
CN101616174A (en) Method for optimizing system performance by dynamically tracking IO processing path of storage system
US20130263142A1 (en) Control device, control method, computer readable recording medium in which program is recorded, and distributed processing system
US20090070648A1 (en) Efficient Scheduling of Background Scrub Commands
US20090313422A1 (en) Flash memory control apparatus having sequential writing procedure and method thereof
US8140804B1 (en) Systems and methods for determining whether to perform a computing operation that is optimized for a specific storage-device-technology type
US8745329B2 (en) Storing data across a plurality of storage nodes
CN102810050A (en) Log data writing method and log system
US20140281280A1 (en) Selecting between non-volatile memory units having different minimum addressable data unit sizes
US20120159096A1 (en) Synchronous extent migration protocol for paired storage
US9052826B2 (en) Selecting storage locations for storing data based on storage location attributes and data usage statistics
US20130198465A1 (en) Connection apparatus, storage apparatus, and computer-readable recording medium having connection request transmission control program recorded therein

Legal Events

Date Code Title Description
C06 Publication
C10 Entry into substantive examination
C14 Grant of patent or utility model