CN106547478A - The I/O throughput monitoring method of process and device - Google Patents

Abstract

The invention discloses a process IO throughput monitoring method and device, and relates to the field of computers. The method includes: obtaining the process identification of each target service; according to each process identification, obtaining the total IO throughput from the IO instance file corresponding to each process identification in the system kernel file every predetermined time; according to the current corresponding to each target service The total IO throughput corresponding to the acquisition time and the total IO throughput of at least one acquisition time before the current acquisition time are calculated to calculate the real-time IO throughput of each target service. The invention can monitor the access situation of each target service to the disk, and can flexibly monitor the respective access situations of the target service for reading and writing to the disk.

Description

Process IO throughput monitoring method and device

technical field

The invention relates to the field of computers, in particular to a process IO throughput monitoring method and device.

Background technique

IOPS (Input/Output Per Second, input and output per second) is one of the main indicators to measure disk performance. IOPS refers to the number of I/O requests that the system can process per unit time. Generally, the number of I/O requests processed per second is taken as the unit. I/O requests are usually read or write data operation requests. IOPS can be understood as IO throughput quantity.

At present, the monitoring of IOPS is all about monitoring the overall IO throughput of the disk, which includes the IO throughput generated by all processes reading and writing to the disk. For example, the IO throughput of the disk obtained by the iostat command is the IO throughput of the disk. It cannot know the IO throughput of the specific service and cannot monitor the target service.

Contents of the invention

In view of the above problems, the present invention is proposed to provide a process IO throughput monitoring device and a corresponding process IO throughput monitoring method that overcome the above problems or at least partially solve the above problems.

According to one aspect of the present invention, a method for monitoring IO throughput of a process is provided, including:

Obtain the process identification of each target service;

According to each process identification, the total IO throughput is obtained from the IO instance file corresponding to each process identification in the system kernel file every predetermined time;

The real-time IO throughput of each target service is calculated according to the total IO throughput corresponding to the current acquisition time corresponding to each target service, and the total IO throughput of at least one acquisition time before the current acquisition time.

Preferably, according to each process identification, the IO total throughput is obtained from the IO instance file corresponding to each process identification in the system kernel file every predetermined time, including:

Obtain the IO instance file corresponding to each process identifier from the system kernel file path /proc/$pid/io every predetermined time; wherein the pid is the process identifier;

From the IO instance file, the total IO throughput at the corresponding moment is extracted.

Preferably, the calculation of the real-time IO throughput of each target service according to the total IO throughput corresponding to the current acquisition moment corresponding to each target service, and the total IO throughput of at least one acquisition moment before the current acquisition moment includes:

For each target service, the total IO throughput corresponding to the current acquisition time is subtracted from the total IO throughput of the previous acquisition time at the current acquisition time to obtain the real-time IO throughput of each target service.

Preferably, the calculation of the real-time IO throughput of each target service according to the total IO throughput corresponding to the current acquisition moment corresponding to each target service, and the total IO throughput of at least one acquisition moment before the current acquisition moment includes:

For each target service, for each acquisition moment, the IO total throughput corresponding to the acquisition moment is subtracted from the IO total throughput at the previous acquisition moment to obtain the first IO throughput;

For the current acquisition moment, at least two recent first IO throughputs from the current acquisition moment are weighted to obtain the real-time IO throughput of each target service.

Preferably, from the IO instance file, the total IO throughput at the corresponding moment is extracted, including:

Extract the total IO read times from the syscr field of the IO instance file, and/or extract the total IO write times from the syscw field of the IO instance file;

The number of IO reads is taken as the total IO throughput, or the number of IO writes is taken as the total IO throughput, or the sum of the number of IO reads and the number of IO writes is taken as the total IO throughput.

Preferably, after calculating the real-time IO throughput of each target service according to the total IO throughput corresponding to the current acquisition moment corresponding to each target service, and the total IO throughput of at least one acquisition moment before the current acquisition moment, it also includes:

Sort each target service according to the real-time IO throughput.

Preferably, after calculating the real-time IO throughput of each target service according to the total IO throughput corresponding to the current acquisition moment corresponding to each target service, and the total IO throughput of at least one acquisition moment before the current acquisition moment, it also includes:

For the real-time IO throughput of each target service, it is judged whether the IO throughput exceeds the threshold; if the IO throughput exceeds the threshold, an alarm interface is called to alarm the target service.

According to another aspect of the present invention, the present invention also discloses a process IO throughput monitoring device, including:

A process identification acquisition module, adapted to obtain the process identification of each target service;

The IO total throughput acquisition module is adapted to obtain the IO total throughput from the IO instance file corresponding to each process identification in the system kernel file at predetermined intervals according to each process identification;

The real-time IO throughput calculation module is adapted to calculate the real-time IO throughput of each target service according to the total IO throughput corresponding to the current acquisition time corresponding to each target service, and the total IO throughput of at least one acquisition time before the current acquisition time .

Preferably, the IO total throughput acquisition module includes:

The IO instance file acquisition submodule is adapted to obtain the IO instance file corresponding to each process identifier from the system kernel file path /proc/$pid/io at predetermined intervals; wherein the pid is the process identifier;

The total IO throughput acquisition sub-module is adapted to extract the total IO throughput at a corresponding moment from the IO instance file.

Preferably, the real-time IO throughput calculation module includes:

The first real-time IO throughput calculation submodule is adapted to subtract the IO total throughput corresponding to the current acquisition moment from the IO total throughput at the previous acquisition moment of the current acquisition moment for each target service to obtain the IO throughput of each target service Real-time IO throughput.

Preferably, the real-time IO throughput calculation module includes:

The independent IO throughput calculation sub-module is suitable for each target service. For each acquisition moment, the total IO throughput corresponding to the acquisition moment is subtracted from the IO total throughput of the acquisition moment before the acquisition moment to obtain The first IO throughput;

The second real-time IO throughput calculation sub-module is adapted to perform weighted calculation on at least two recent first IO throughputs from the current acquisition moment to obtain the real-time IO throughput of each target service.

Preferably, the IO total throughput acquisition submodule includes:

The times extraction module is adapted to extract the total IO read times from the syscr field of the IO instance file, and/or extract the total IO write times from the syscw field of the IO instance file;

The throughput confirmation module is adapted to use the IO read times as the total IO throughput, or use the IO write times as the IO total throughput, or use the sum of the IO read times and the IO write times as the IO total throughput .

Preferably, after the real-time IO throughput calculation module, it also includes:

The sorting module is adapted to sort each target service according to the real-time IO throughput.

Preferably, after the real-time IO throughput calculation module, it also includes:

The alarm module is adapted to judge whether the IO throughput exceeds a threshold for the real-time IO throughput of each target service; if the IO throughput exceeds the threshold, call an alarm interface to alarm the target service.

According to the IO throughput monitoring method and device of the process of the present invention, the process identification of each target service can be obtained, and according to each process identification, every predetermined time, the IO total is obtained from the IO instance file corresponding to each process identification in the system kernel file. Throughput, and then calculate the real-time IO throughput of each target service according to the total IO throughput corresponding to the current acquisition time corresponding to each target service, and the total IO throughput of at least one acquisition time before the current acquisition time, for example, for service A , the total IO throughput at the current acquisition time is 10000 reads, and the total IO throughput at the previous acquisition time is 9500 reads, then the real-time IO throughput corresponding to the current acquisition time can be 10000-9500=500 times, by This solves the problem that the prior technology cannot target a specific target service and monitor the real-time IO throughput of the service, and achieves the ability to monitor the real-time IO throughput of a specific target service to facilitate subsequent monitoring of the target service. Beneficial effects of control.

The above description is only an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention, it can be implemented according to the contents of the description, and in order to make the above and other purposes, features and advantages of the present invention more obvious and understandable , the specific embodiments of the present invention are enumerated below.

Description of drawings

Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiment. The drawings are only for the purpose of illustrating a preferred embodiment and are not to be considered as limiting the invention. Also throughout the drawings, the same reference numerals are used to designate the same parts. In the attached picture:

FIG. 1 shows a schematic flow diagram of a process IO throughput monitoring method according to an embodiment of the present invention;

FIG. 2 shows a schematic flow diagram of a process IO throughput monitoring method according to another embodiment of the present invention;

FIG. 3 shows a schematic flow diagram of a process IO throughput monitoring method according to another embodiment of the present invention;

FIG. 4 shows a schematic flow diagram of a process IO throughput monitoring method according to another embodiment of the present invention;

FIG. 5 shows a schematic structural diagram of an apparatus for monitoring IO throughput of a process according to an embodiment of the present invention;

FIG. 6 shows a schematic structural diagram of an apparatus for monitoring IO throughput of a process according to another embodiment of the present invention;

FIG. 7 shows a schematic structural diagram of an apparatus for monitoring IO throughput of a process according to another embodiment of the present invention;

Fig. 8 shows a schematic structural diagram of an apparatus for monitoring IO throughput of a process according to another embodiment of the present invention.

detailed description

Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided for more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

Embodiment one

Referring to FIG. 1 , it shows a schematic flow diagram of a process IO throughput monitoring method, which may specifically include:

Step 110, obtain the process identification of each target service.

For a computer system, a service itself is a process. However, in the embodiment of the present invention, multiple services may be set in one server to handle requests from different applications and provide support for different businesses. It can be understood that the server of the present invention is a single-machine multi-entity server, and each entity is a Serve.

In the embodiment of the present invention, the process identification of each target service started in the server can be obtained, and the process identification is pid in practical application.

Step 120, according to each process ID, obtain the total IO throughput from the IO instance file corresponding to each process ID in the system kernel file every predetermined time.

In practical applications, for the IO throughput of the disk in systems such as the Linux system, from the time the process starts, as time goes by, the IO instance file in the system kernel file can always record the read/write data of the process in a cumulative manner. I/O throughput. For example, for the target service A, its startup time is 12:00:00, then every time a disk read or write operation is performed, the number of times will be accumulated in the corresponding IO instance file.

The aforementioned system kernel files are such as the /proc file system, in which there are IO instance files that record information related to the IO throughput of each target service. The /proc filesystem is a mechanism used by the kernel and kernel modules to send information to processes. This pseudo-filesystem allows applications to interact with kernel internal data structures, obtain useful information about processes, and change settings (by changing kernel parameters) on the fly. Unlike other file systems, /proc exists in memory rather than on the hard disk. /proc is controlled by the kernel, there is no device hosting /proc. Because /proc mainly stores state information controlled by the kernel, the logical location of most of this information is in kernel-controlled memory.

Of course, in the embodiment of the present invention, the IO instance file identified by the process corresponding to each target service may be selected from the above-mentioned system kernel file at predetermined intervals, such as every second. Of course, the predetermined time of the interval can be set to other values, such as 2 seconds, etc., which is not limited by the present invention.

Of course, in the embodiment of the present invention, before step 110, the root authority of the system needs to be obtained first, and the root authority can access any file system of the system. Because the system kernel files are transparent to users with ordinary permissions, users cannot operate these files, and thus cannot read these files from the memory and obtain information. Taking the linux system as an example, you can use the su command to obtain system permissions, such as

[beinan@localhost～]$su

Password:

[root@localhost beinan]#pwd

/home/beinan

The pseudo-code indicates that it switches to the root user by default, and changes to the root user's environment.

Step 130: Calculate the real-time IO throughput of each target service according to the total IO throughput corresponding to the current acquisition time of each target service and the total IO throughput of at least one acquisition time before the current acquisition time.

It can be understood that the above-mentioned every second can be understood as an acquisition time. For example, after the IO throughput monitoring step of the process of the present invention is executed, obtain the total IO throughput in the IO instance file of each process identification at 10:00:00 for the first time, and obtain the total IO throughput at 10:00:01 second The total IO throughput is obtained from the IO instance file for which each process ID is obtained for the first time, and the total IO throughput is obtained from the IO instance file for which each process ID is obtained for the third time at 10:00:02, then the above-mentioned interval of 1 second is The moment is the acquisition moment, which can be understood as the monitoring moment.

Certainly, the total IO throughput is obtained from the IO instance file of each process identifier obtained for the first time, and step 130 does not yet calculate the real-time IO throughput of each target service at this time. After the total IO throughput is obtained from the IO instance file of each process identifier obtained for the second time, step 130 can calculate the real-time IO throughput of each target service.

For the monitoring of the process, in order to ensure its real-time performance, the embodiment of the present invention is based on the IO total throughput in the latest acquired IO instance file, combined with the latest or recent multiple aforementioned IO total throughput, Calculate the total real-time IO throughput corresponding to the current acquisition moment. In the embodiment of the present invention, the total IO throughput can be understood as the total number of times the target service reads and writes the disk, or it can also be understood as the total number of times the target service reads the disk, or it can also be understood as the number of times the target service reads the disk. The total number of writes.

For example, in the foregoing example, the total number of readings of the target service A obtained at 10:00:00 is 100, and the total number of readings of the target service A obtained at 10:00:01 is 200. Assuming that the current acquisition time and the latest acquisition time are used as calculation objects, then for 10:00:01, the number of real-time reads corresponding to the target service A is 200-100=100. Further, the total number of reads of the target service A obtained at 10:00:02 is 400, and the corresponding real-time reads of the target service A are 400-200=200 times.

Assuming the current acquisition time and multiple times before it, you can subtract the previous total IO throughput from the next total IO throughput in order to get multiple values, and then calculate the average based on these values as the real-time IO throughput at the current moment. For example, the total IO throughput at 10:00:01 minus the total IO throughput at 10:00:00 is 100; the total IO throughput at 10:00:02 minus the total IO throughput at 10:00:01 The amount is 200, then the real-time IO throughput corresponding to 10:00:02 is (100+200)/2=150 times.

This solves the problem that the prior technology cannot target a specific target service and cannot monitor the real-time IO throughput of the service, and achieves the ability to monitor the real-time IO throughput of a specific target service to facilitate subsequent services to the target Beneficial effects of control.

In the embodiment of the present invention, technicians have discovered that the system kernel file can record the number of accesses of all processes in the process of accessing the disk in a cumulative form, so the embodiment of the present invention can first find the process ID of the target service, and then use the process ID to Find the IO instance file corresponding to each process ID in the system kernel file to obtain the total IO throughput, and then after obtaining the above total IO throughput each time, you can combine the previous record to calculate the real-time IO throughput of each target service quantity. Therefore, the access status of each target service to the disk can be monitored, avoiding the problem in the prior art that only the disk can be used as the main body to monitor the overall access status of the disk from the outside world.

Embodiment two

Referring to FIG. 2, it shows a schematic flow diagram of a process IO throughput monitoring method, which may specifically include:

Step 210, obtaining the process identification of each target service;

In the embodiment of the present invention, the process identifier of the target service can be obtained through proc/$pid/cmdline. The present invention can traverse the process-related information recorded in the file of proc/$pid/cmdline whose root directory is proc/$pid/cmdline in the proc file system, wherein pid is the process identifier, and the process related information of the pid is recorded in the proc/$pid/cmdline file Information, such as whether it is a child process, if it is a child process, the process ID of its parent process, the process handle corresponding to the pid, and other process-related information. Then compare the obtained process information with the target service to confirm the process identifier pid of the target service process.

Step 220: Obtain the IO instance file corresponding to each process ID from the system kernel file path /proc/$pid/io every predetermined time; wherein the pid is the process ID.

In the embodiment of the present invention, the system kernel file, that is, the IO instance file recording each target service can be accessed according to its root directory /proc/$pid/io in the /proc file system. For example, if the pid of the target service is 719116688, then the path of its IO instance file is /proc/$19573/io.

Of course, in practical applications, there is a root directory of the system before /proc, such as a Linux system, its complete path can be "root directory/proc/$19573/io", and the specific form of the root directory depends on the system situation. Then you can read the IO instance file of 19573 from the "root directory/proc/$19573/io" path.

Step 230, extract the total IO throughput at the corresponding moment from the IO instance file.

In the embodiment of the present invention, in the IO instance file of the system kernel file, the number of times the process accesses the disk is accumulated to obtain the total IO throughput of each target service at the time of its acquisition.

Preferably, from the IO instance file, the total IO throughput at the corresponding moment is extracted, including:

Substep 231, extracting the total IO read times from the syscr field of the IO instance file. And/or extracting the total IO write times from the syscw field of the IO instance file;

In sub-step 232, the number of IO reads is taken as the total IO throughput, or the number of IO writes is taken as the total IO throughput, or the sum of the number of IO reads and the number of IO writes is taken as the total IO throughput.

In practical applications, the content of the instance file corresponding to /proc/19573/io is as follows:

rchar:867994155

wchar:2352795453

syscr:243617

syscw:1299355

read_bytes: 171048960

write_bytes: 2199875584

canceled_write_bytes: 51793920

Among them, rchar is the total number of bytes read from the disk, the sum of the length parameters in read or pread, which is calculated from the pagecache (page cache), and does not represent the actual disk read; wchar is written to the disk The total number of bytes, the sum of the length parameters in write or pwrite; syscr is the total number of reads to the disk; syscr is the total number of writes to the disk; read_bytes is the total number of bytes actually read from the disk; write_bytes is the actual The total number of bytes written to disk; canceled_write_bytes is the number of bytes written that should have occurred but did not occur due to truncated pagecache.

Then, in order to count the total IO throughput, the present invention can take the reading and/or writing of the disk as its total IO throughput. As in the above example, the total IO read count 243617 can be extracted from the syscr field of the IO instance file, and/or the total IO write count 1299355 can be extracted from the syscw field of the IO instance file. Then 1299355+243617 can be used as the total IO throughput of the corresponding target service, or 243617 can be used alone as the IO throughput of the corresponding target service, or 1299355 can be used alone as the IO throughput of the corresponding target service. This enables the present invention to more flexibly monitor the number of reads and writes to the disk by the target service.

Step 240, for each target service, subtract the total IO throughput at the previous acquisition time from the total IO throughput corresponding to the current acquisition time to obtain the real-time IO throughput of each target service.

For example, the aforementioned IO instance file of 19573 was obtained at 10:00:00, from which syscr:243617 and syscw:1299355 were extracted. At 10:00:01, the aforementioned instance file of 19573 was obtained again, and syscr:243700 and syscw:1299800 were extracted from it. Then corresponding to the time 10:00:01, the number of real-time reads of the target service of 19573 is (243700-243617)=83, the number of real-time writes is (1299800-1299355)=445, and the number of real-time read+writes is 528.

Of course, in the embodiment of the present invention, if the predetermined time interval is greater than 1 second, the result of the above subtraction can be divided by the predetermined time to obtain the real-time IO throughput per second.

In the embodiment of the present invention, technicians have discovered that the system kernel file can record the number of accesses of all processes in the process of accessing the disk in a cumulative form, so the embodiment of the present invention can first find the process ID of the target service, and then use the process ID to Find the IO instance file corresponding to each process ID in the system kernel file to obtain the total IO throughput, and then after obtaining the above total IO throughput each time, you can combine the previous record to calculate the real-time IO throughput of each target service quantity. Therefore, the access status of each target service to the disk can be monitored, and the respective access status of the target service to the disk for reading and writing can be flexibly monitored.

Embodiment Three

Referring to FIG. 3 , it shows a schematic flow diagram of a process IO throughput monitoring method, which may specifically include:

Step 310, obtaining the process identification of each target service;

Step 320, at predetermined intervals, obtain the IO instance file corresponding to each process ID from the system kernel file path /proc/$pid/io; wherein the pid is the process ID.

Step 330, extract the total IO throughput at the corresponding moment from the IO instance file.

Step 340, for each target service, for each acquisition time, subtract the total IO throughput at the acquisition time preceding the acquisition time from the total IO throughput corresponding to the acquisition time, to obtain the first IO throughput;

Step 350 , for the current acquisition moment, perform weighted calculation on at least two recent first IO throughputs from the current acquisition moment, so as to obtain the real-time IO throughput of each target service.

For example, the aforementioned IO instance file of 19573 was obtained at 10:00:00, from which syscr:243617 and syscw:1299355 were extracted. At 10:00:01, the aforementioned instance file of 19573 was obtained again, and syscr:243700 and syscw:1299800 were extracted from it. At 10:00:02, the aforementioned instance file of 19573 was obtained again, from which syscr:243800 and syscw:1299900 were extracted.

Then the data subtracting 10:00:00 from 10:00:01 is: the number of reads of the target service of 19573 is (243700-243617) = 83, the number of writes is (1299800-1299355) = 445, read + write The number of times is 528.

Then the data minus 10:00:01 from 10:00:02 is: the number of reads of the target service of 19573 is (243800-243700) = 100, the number of writes is (1299900-1299800) = 100, read + write The number of times is 200.

Then for the current acquisition time 10:00:02, the number of real-time reads can be (100+83)/2, the number of real-time writes can be (100+445)/2, and the number of real-time read+writes can be (200+528) /2.

Of course, the closer the record is to the current acquisition time, the higher its weight can be set. For example, the weight of the data subtracted from 10:00:02 by 10:00:01 is set to 0.7, and the weight of the data subtracted from 10:00:01 by 10:00:00 is set to 0.3, then real-time read + write For example, the number of times is 200*0.7+528*0.3.

Of course, in the embodiment of the present invention, if the predetermined time interval is greater than 1 second, the result of the above subtraction can be divided by the predetermined time to obtain the real-time IO throughput per second.

In the embodiment of the present invention, technicians have discovered that the system kernel file can record the number of accesses of all processes in the process of accessing the disk in a cumulative form, so the embodiment of the present invention can first find the process ID of the target service, and then use the process ID to Find the IO instance file corresponding to each process ID in the system kernel file to obtain the total IO throughput, and then after obtaining the above total IO throughput each time, you can combine the previous record to calculate the real-time IO throughput of each target service quantity. In this way, the access status of each target service to the disk can be monitored, and the access status of the target service to the disk for reading and writing can be flexibly monitored. In addition, the real-time IO throughput at the current moment can be smoothed through multiple records data, making it less abnormal.

Embodiment Four

Referring to FIG. 4 , it shows a schematic flow chart of a process IO throughput monitoring method, which may specifically include:

Step 410, obtain the process identification of each target service.

Step 420, according to each process ID, obtain the total IO throughput from the IO instance file corresponding to each process ID in the system kernel file every predetermined time.

Step 430: Calculate the real-time IO throughput of each target service according to the total IO throughput corresponding to the current acquisition time of each target service and the total IO throughput of at least one acquisition time before the current acquisition time.

Step 440, sort each target service according to the real-time IO throughput.

In the embodiment of the present invention, for each target service, at the same acquisition moment, they can be sorted according to their real-time IO throughput, so that it is convenient to judge which ones may exceed the threshold and which ones do not.

Step 450, for the real-time IO throughput of each target service, judge whether the IO throughput exceeds a threshold; if the IO throughput exceeds the threshold, call an alarm interface to give an alarm to the target service.

In the embodiment of the present invention, after monitoring the number of disk accesses of the target service, a threshold for real-time IO throughput and an alarm interface will be preset in order to facilitate technicians to handle services that may affect server performance in a timely manner , the alarm interface can send alarm information to technicians, such as sending alarm emails to technician mailboxes, sending text messages to technicians’ mobile phones, etc. At each acquisition moment, compare the obtained real-time IO throughput of each target service with the threshold. If it exceeds, this indicates that the target service may have a problem and occupy too many resources, and an alarm needs to be issued. The alarm information can include The historical access data of the target process to the disk, and the real-time IO throughput at each acquisition time obtained by the present invention.

Of course, while calling the police, the monitoring is also continued.

If the threshold is not exceeded, no alarm will be issued and monitoring will continue.

Of course, in the embodiment of the present invention, after judging whether the IO throughput exceeds a threshold for the real-time IO throughput of each target service, operations such as temporarily reducing the processing volume of the target service for requests can also be performed. The present invention is not limited thereto.

In the embodiment of the present invention, technicians have discovered that the system kernel file can record the number of accesses of all processes in the process of accessing the disk in a cumulative form, so the embodiment of the present invention can first find the process ID of the target service, and then use the process ID to Find the IO instance file corresponding to each process ID in the system kernel file to obtain the total IO throughput, and then after obtaining the above total IO throughput each time, you can combine the previous record to calculate the real-time IO throughput of each target service quantity. In this way, the access status of each target service to the disk can be monitored, and the access status of the target service to the disk for reading and writing can be flexibly monitored; in addition, the real-time IO throughput at the current moment can be smoothed through multiple records The data makes it less abnormal; moreover, the target service whose real-time IO throughput exceeds the threshold can generate an alarm message to notify the technician, so that the technician can know in time which target service occupies a particularly large IO. So that it can be dealt with in a targeted manner.

Embodiment five

Referring to FIG. 5, it shows a schematic structural diagram of a process IO throughput monitoring device, which may specifically include:

The process identification obtaining module 510 is adapted to obtain the process identification of each target service;

The IO total throughput acquisition module 520 is adapted to obtain the IO total throughput from the IO instance file corresponding to each process identification in the system kernel file at regular intervals according to each process identification;

The real-time IO throughput calculation module 530 is adapted to calculate the real-time IO throughput of each target service according to the total IO throughput corresponding to the current acquisition time corresponding to each target service, and the total IO throughput of at least one acquisition time before the current acquisition time quantity.

In the embodiment of the present invention, technicians have discovered that the system kernel file can record the number of accesses of all processes in the process of accessing the disk in a cumulative form, so the embodiment of the present invention can first find the process ID of the target service, and then use the process ID to Find the IO instance file corresponding to each process ID in the system kernel file to obtain the total IO throughput, and then after obtaining the above total IO throughput each time, you can combine the previous record to calculate the real-time IO throughput of each target service quantity. Therefore, the access status of each target service to the disk can be monitored, avoiding the problem in the prior art that only the disk can be used as the main body to monitor the overall access status of the disk from the outside world.

Embodiment six

Referring to FIG. 6, it shows a schematic structural diagram of a process IO throughput monitoring device, which may specifically include:

The process identification obtaining module 610 is adapted to obtain the process identification of each target service;

The IO total throughput acquisition module 620 is adapted to obtain the IO total throughput from the IO instance file corresponding to each process identification in the system kernel file at regular intervals according to each process identification, specifically including:

The IO instance file acquisition sub-module 621 is adapted to acquire the IO instance file corresponding to each process ID from the system kernel file path /proc/$pid/io every predetermined time; wherein the pid is the process ID.

The total IO throughput acquisition sub-module 622 is adapted to extract the total IO throughput at a corresponding moment from the IO instance file.

The real-time IO throughput calculation module 630 is adapted to calculate the real-time IO throughput of each target service according to the total IO throughput corresponding to the current acquisition time corresponding to each target service, and the total IO throughput of at least one acquisition time before the current acquisition time amount, including:

The first real-time IO throughput calculation sub-module 631 is adapted to subtract the total IO throughput at the previous acquisition moment from the total IO throughput at the current acquisition moment for each target service to obtain each target service real-time IO throughput.

Preferably, the IO total throughput acquisition submodule includes:

The times extraction module is adapted to extract the total IO read times from the syscr field of the IO instance file, and/or extract the total IO write times from the syscw field of the IO instance file;

The throughput confirmation module is adapted to use the IO read times as the total IO throughput, or use the IO write times as the IO total throughput, or use the sum of the IO read times and the IO write times as the IO total throughput .

In the embodiment of the present invention, technicians have discovered that the system kernel file can record the number of accesses of all processes in the process of accessing the disk in a cumulative form, so the embodiment of the present invention can first find the process ID of the target service, and then use the process ID to Find the IO instance file corresponding to each process ID in the system kernel file to obtain the total IO throughput, and then after obtaining the above total IO throughput each time, you can combine the previous record to calculate the real-time IO throughput of each target service quantity. Therefore, the access status of each target service to the disk can be monitored, and the respective access status of the target service to the disk for reading and writing can be flexibly monitored.

Embodiment seven

Referring to FIG. 7 , it shows a schematic structural diagram of a process IO throughput monitoring device, which may specifically include:

The process identification obtaining module 710 is adapted to obtain the process identification of each target service;

The IO total throughput acquisition module 720 is adapted to obtain the IO total throughput from the IO instance file corresponding to each process identification in the system kernel file at regular intervals according to each process identification, specifically including:

The IO instance file acquisition sub-module 721 is adapted to acquire the IO instance file corresponding to each process ID from the system kernel file path /proc/$pid/io at predetermined intervals; wherein the pid is the process ID.

The total IO throughput acquisition sub-module 722 is adapted to extract the total IO throughput at a corresponding moment from the IO instance file.

The real-time IO throughput calculation module 730 is adapted to calculate the real-time IO throughput of each target service according to the total IO throughput corresponding to the current acquisition time corresponding to each target service, and the total IO throughput of at least one acquisition time before the current acquisition time amount, including:

The independent IO throughput calculation sub-module 731 is suitable for each target service, and for each acquisition moment, subtracting the total IO throughput corresponding to the acquisition moment from the total IO throughput at the acquisition moment before the acquisition moment, Get the first IO throughput;

The second real-time IO throughput calculation sub-module 732 is adapted to perform weighted calculation on at least two recent first IO throughputs from the current acquisition moment to obtain the real-time IO throughput of each target service.

In the embodiment of the present invention, technicians have discovered that the system kernel file can record the number of accesses of all processes in the process of accessing the disk in a cumulative form, so the embodiment of the present invention can first find the process ID of the target service, and then use the process ID to Find the IO instance file corresponding to each process ID in the system kernel file to obtain the total IO throughput, and then after obtaining the above total IO throughput each time, you can combine the previous record to calculate the real-time IO throughput of each target service quantity. In this way, the access status of each target service to the disk can be monitored, and the access status of the target service to the disk for reading and writing can be flexibly monitored. In addition, the real-time IO throughput at the current moment can be smoothed through multiple records data, making it less abnormal.

Embodiment eight

Referring to FIG. 8 , it shows a schematic structural diagram of a process IO throughput monitoring device, which may specifically include:

A process identification obtaining module 810, adapted to obtain the process identification of each target service;

IO total throughput acquisition module 820 is adapted to obtain the total IO throughput from the IO instance file corresponding to each process identification in the system kernel file at regular intervals according to each process identification;

The real-time IO throughput calculation module 830 is adapted to calculate the real-time IO throughput of each target service according to the total IO throughput corresponding to the current acquisition time corresponding to each target service, and the total IO throughput of at least one acquisition time before the current acquisition time quantity.

The sorting module 840 is adapted to sort each target service according to the real-time IO throughput.

The alarm module 850 is adapted to judge whether the IO throughput exceeds a threshold for the real-time IO throughput of each target service; if the IO throughput exceeds the threshold, call an alarm interface to alarm the target service.

In the embodiment of the present invention, technicians have discovered that the system kernel file can record the number of accesses of all processes in the process of accessing the disk in a cumulative form, so the embodiment of the present invention can first find the process ID of the target service, and then use the process ID to Find the IO instance file corresponding to each process ID in the system kernel file to obtain the total IO throughput, and then after obtaining the above total IO throughput each time, you can combine the previous record to calculate the real-time IO throughput of each target service quantity. In this way, the access status of each target service to the disk can be monitored, and the access status of the target service to the disk for reading and writing can be flexibly monitored; in addition, the real-time IO throughput at the current moment can be smoothed through multiple records The data makes it less abnormal; moreover, the target service whose real-time IO throughput exceeds the threshold can generate an alarm message to notify the technician, so that the technician can know in time which target service occupies a particularly large IO. So that it can be dealt with in a targeted manner.

The algorithms and displays presented herein are not inherently related to any particular computer, virtual system, or other device. Various generic systems can also be used with the teachings based on this. The structure required to construct such a system is apparent from the above description. Furthermore, the present invention is not specific to any particular programming language. It should be understood that various programming languages can be used to implement the content of the present invention described herein, and the above description of specific languages is for disclosing the best mode of the present invention.

In the description provided herein, numerous specific details are set forth. However, it is understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure the understanding of this description.

Similarly, it should be appreciated that in the foregoing description of exemplary embodiments of the invention, in order to streamline this disclosure and to facilitate an understanding of one or more of the various inventive aspects, various features of the invention are sometimes grouped together in a single embodiment, figure, or its description. This method of disclosure, however, is not to be interpreted as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed embodiment. Thus, the claims following the Detailed Description are hereby expressly incorporated into this Detailed Description, with each claim standing on its own as a separate embodiment of this invention.

Those skilled in the art can understand that the modules in the device in the embodiment can be adaptively changed and arranged in one or more devices different from the embodiment. Modules or units or components in the embodiments may be combined into one module or unit or component, and furthermore may be divided into a plurality of sub-modules or sub-units or sub-assemblies. All features disclosed in this specification (including accompanying claims, abstract and drawings) and any method or method so disclosed may be used in any combination, except that at least some of such features and/or processes or units are mutually exclusive. All processes or units of equipment are combined. Each feature disclosed in this specification (including accompanying claims, abstract and drawings) may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise.

Furthermore, those skilled in the art will understand that although some embodiments described herein include some features included in other embodiments but not others, combinations of features from different embodiments are meant to be within the scope of the invention. and form different embodiments. For example, in the following claims, any of the claimed embodiments may be used in any combination.

The various component embodiments of the present invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art should understand that a microprocessor or a digital signal processor (DSP) can be used in practice to realize some or all functions of some or all of the components in the IO throughput monitoring device of the process according to the embodiment of the present invention . The present invention can also be implemented as an apparatus or an apparatus program (for example, a computer program and a computer program product) for performing a part or all of the methods described herein. Such a program for realizing the present invention may be stored on a computer-readable medium, or may be in the form of one or more signals. Such a signal may be downloaded from an Internet site, or provided on a carrier signal, or provided in any other form.

It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention can be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means can be embodied by one and the same item of hardware. The use of the words first, second, and third, etc. does not indicate any order. These words can be interpreted as names.

The invention discloses A1, a process IO throughput monitoring method, comprising:

Obtain the process identification of each target service;

According to each process identification, the total IO throughput is obtained from the IO instance file corresponding to each process identification in the system kernel file every predetermined time;

The real-time IO throughput of each target service is calculated according to the total IO throughput corresponding to the current acquisition time corresponding to each target service, and the total IO throughput of at least one acquisition time before the current acquisition time.

A2, according to the method described in A1, described according to each process identification, every predetermined time, obtains IO total throughput from the IO instance file corresponding to each process identification in the system kernel file, including:

Obtain the IO instance file corresponding to each process identifier from the system kernel file path /proc/$pid/io every predetermined time; wherein the pid is the process identifier;

From the IO instance file, the total IO throughput at the corresponding moment is extracted.

A3. According to the method described in A1 or A2, according to the total IO throughput corresponding to the current acquisition moment corresponding to each target service, and the total IO throughput of at least one acquisition moment before the current acquisition moment, calculate the IO throughput of each target service Real-time IO throughput, including:

For each target service, the total IO throughput corresponding to the current acquisition time is subtracted from the total IO throughput of the previous acquisition time at the current acquisition time to obtain the real-time IO throughput of each target service.

A4. According to the method described in A1 or A2, according to the total IO throughput corresponding to the current acquisition moment corresponding to each target service, and the total IO throughput of at least one acquisition moment before the current acquisition moment, calculate the IO throughput of each target service Real-time IO throughput, including:

For each target service, for each acquisition moment, the IO total throughput corresponding to the acquisition moment is subtracted from the IO total throughput at the previous acquisition moment to obtain the first IO throughput;

For the current acquisition moment, at least two recent first IO throughputs from the current acquisition moment are weighted to obtain the real-time IO throughput of each target service.

A5. According to the method described in A2, from the IO instance file, extract the total IO throughput at the corresponding moment, including:

Extract the total IO read times from the syscr field of the IO instance file, and/or extract the total IO write times from the syscw field of the IO instance file;

The number of IO reads is taken as the total IO throughput, or the number of IO writes is taken as the total IO throughput, or the sum of the number of IO reads and the number of IO writes is taken as the total IO throughput.

A6. According to the method described in A1, the real-time IO throughput of each target service is calculated according to the total IO throughput corresponding to the current acquisition time corresponding to each target service, and the total IO throughput of at least one acquisition time before the current acquisition time After the amount, also include:

Sort each target service according to the real-time IO throughput.

A7. According to the method described in A1, calculate the real-time IO throughput of each target service according to the total IO throughput corresponding to the current acquisition time corresponding to each target service, and the total IO throughput of at least one acquisition time before the current acquisition time After the amount, also include:

For the real-time IO throughput of each target service, it is judged whether the IO throughput exceeds the threshold; if the IO throughput exceeds the threshold, an alarm interface is called to alarm the target service.

The present invention also discloses B8, a process IO throughput monitoring device, comprising:

A process identification acquisition module, adapted to obtain the process identification of each target service;

The IO total throughput acquisition module is adapted to obtain the IO total throughput from the IO instance file corresponding to each process identification in the system kernel file at predetermined intervals according to each process identification;

The real-time IO throughput calculation module is adapted to calculate the real-time IO throughput of each target service according to the total IO throughput corresponding to the current acquisition time corresponding to each target service, and the total IO throughput of at least one acquisition time before the current acquisition time .

B9, according to the device described in B8, the IO total throughput acquisition module includes:

The IO instance file acquisition submodule is adapted to obtain the IO instance file corresponding to each process identifier from the system kernel file path /proc/$pid/io at predetermined intervals; wherein the pid is the process identifier;

The total IO throughput acquisition sub-module is adapted to extract the total IO throughput at a corresponding moment from the IO instance file.

B10, according to the device described in B8 or B9, the real-time IO throughput calculation module includes:

The first real-time IO throughput calculation submodule is adapted to subtract the IO total throughput corresponding to the current acquisition moment from the IO total throughput at the previous acquisition moment of the current acquisition moment for each target service to obtain the IO throughput of each target service Real-time IO throughput.

B11, according to the device described in B8 or B9, the real-time IO throughput calculation module includes:

The independent IO throughput calculation sub-module is suitable for each target service. For each acquisition moment, the total IO throughput corresponding to the acquisition moment is subtracted from the IO total throughput of the acquisition moment before the acquisition moment to obtain The first IO throughput;

The second real-time IO throughput calculation sub-module is adapted to perform weighted calculation on at least two recent first IO throughputs from the current acquisition moment to obtain the real-time IO throughput of each target service.

B12, according to the device described in B9, the IO total throughput acquisition submodule includes:

The times extraction module is adapted to extract the total IO read times from the syscr field of the IO instance file, and/or extract the total IO write times from the syscw field of the IO instance file;

The throughput confirmation module is adapted to use the IO read times as the total IO throughput, or use the IO write times as the IO total throughput, or use the sum of the IO read times and the IO write times as the IO total throughput .

B13, according to the device described in B8, after the real-time IO throughput calculation module, it also includes:

The sorting module is adapted to sort each target service according to the real-time IO throughput.

B14, according to the device described in B8, after the real-time IO throughput calculation module, it also includes:

The alarm module is adapted to judge whether the IO throughput exceeds a threshold for the real-time IO throughput of each target service; if the IO throughput exceeds the threshold, call an alarm interface to alarm the target service.

Claims (10)

1. the I/O throughput monitoring method of a kind of process, including：

Obtain the process identification (PID) of each destination service；

According to each process identification (PID), at predetermined time intervals, each process identification (PID) of correspondence from system kernel file IO total throughouts are obtained in IO instance documents；

According to the corresponding IO total throughouts of current acquisition moment for corresponding to each destination service, and currently The IO total throughouts at least one acquisition moment before the moment are obtained, the reality of each destination service is calculated When I/O throughput.

2. method according to claim 1, it is characterised in that described according to each process identification (PID), At predetermined time intervals, IO is obtained in the IO instance documents of each process identification (PID) of correspondence from system kernel file Total throughout, including：

At predetermined time intervals, each process mark of acquisition correspondence from system kernel file path/proc/ $ pid/io Know corresponding IO instance documents；Wherein described pid is process identification (PID)；

From the IO instance documents, the IO total throughouts at corresponding moment are extracted.

3. method according to claim 1 and 2, it is characterised in that it is described according to correspondence each Before the corresponding IO total throughouts of current acquisition moment of destination service, and current acquisition moment extremely The IO total throughouts at a few acquisition moment, calculate the real-time I/O throughput of each destination service, including：

For each destination service, moment corresponding IO total throughouts will be currently obtained, is deducted and is currently obtained The IO total throughouts at the previous acquisition moment at moment are taken, is gulped down with the real-time IO for obtaining each destination service The amount of telling.

4. method according to claim 1 and 2, the working as according to each destination service of correspondence It is front to obtain moment corresponding IO total throughouts, and when currently obtaining at least one acquisition before the moment The IO total throughouts at quarter, calculate the real-time I/O throughput of each destination service, including：

For each destination service, for each acquisition moment, will be corresponding at acquisition moment IO total Handling capacity, deducts the IO total throughouts for obtaining the previous acquisition moment at moment, obtains an IO and gulp down The amount of telling；

The moment is obtained for current, will be from the current acquisition moment, at least two the oneth nearest IO are gulped down The amount of telling is weighted, to obtain the real-time I/O throughput of each destination service.

5. method according to claim 2, it is characterised in that from the IO instance documents, The IO total throughouts at corresponding moment are extracted, including：

Total IO is extracted from the syscr fields of the IO instance documents and reads number of times, and/or from the IO Total IO is extracted in the syscw fields of instance document and writes number of times；

The IO is read number of times as IO total throughouts, or the IO is write number of times always to gulp down as IO The amount of telling, or IO reading number of times is write into number of times sum as IO total throughouts with IO.

6. method according to claim 1, it is characterised in that take in each target according to correspondence Before the corresponding IO total throughouts of current acquisition moment of business, and current acquisition moment at least one The IO total throughouts at moment are obtained, after calculating the real-time I/O throughput of each destination service, is also included：

Each destination service is ranked up according to real-time I/O throughput size.

7. method according to claim 1, it is characterised in that take in each target according to correspondence Before the corresponding IO total throughouts of current acquisition moment of business, and current acquisition moment at least one The IO total throughouts at moment are obtained, after calculating the real-time I/O throughput of each destination service, is also included：

For the real-time I/O throughput of each destination service, judge whether the I/O throughput exceedes threshold value； If the I/O throughput exceedes threshold value, remote alarm interface is called to report to the police the destination service.

8. the I/O throughput supervising device of a kind of process, including：

Process identification (PID) acquisition module, is suitable to obtain the process identification (PID) of each destination service；

IO total throughout acquisition modules, are suitable to according to each process identification (PID), at predetermined time intervals, from system IO total throughouts are obtained in the IO instance documents of each process identification (PID) of correspondence in kernel file；

I/O throughput computing module, is suitable to the current acquisition moment according to each destination service of correspondence in real time The IO at least one acquisition moment before corresponding IO total throughouts, and current acquisition moment is always gulped down The amount of telling, calculates the real-time I/O throughput of each destination service.

9. device according to claim 8, it is characterised in that the IO total throughouts obtain mould Block, including：

IO instance document acquisition submodules, are suitable at predetermined time intervals, from system kernel file path The corresponding IO instance documents of each process identification (PID) of correspondence are obtained in/proc/ $ pid/io；Wherein described pid be into Journey is identified；

IO total throughout acquisition submodules, are suitable to from the IO instance documents, extract the corresponding moment IO total throughouts.

10. device according to claim 8 or claim 9, it is characterised in that the real-time IO handles up Amount computing module, including：

First real-time I/O throughput calculating sub module, is suitable to, for each destination service, currently to obtain Moment corresponding IO total throughouts, the IO for deducting the current previous acquisition moment for obtaining the moment are always handled up Amount, to obtain the real-time I/O throughput of each destination service.