CN111953544B - Fault detection method, device, equipment and storage medium of server - Google Patents

Abstract

This application discloses a server fault detection method, including: collecting fault data information generated when the server fails; performing word segmentation processing on each fault data information to obtain corresponding keywords; determining each keyword and the preset word segmentation category The matching times of the preset keywords in the library; the fault value of the server is calculated according to the matched keywords and the corresponding matching times, and the fault condition of the server is determined according to the fault value. The method can improve the convenience of fault detection for servers of different manufacturers or servers of different operating systems, thereby improving the efficiency of fault detection for servers. The application also discloses a server fault detection device, equipment and computer-readable storage medium, all of which have the above-mentioned beneficial effects.

Description

A server fault detection method, device, equipment and storage medium

technical field

The invention relates to the field of server detection, in particular to a server fault detection method, device, equipment and computer-readable storage medium.

Background technique

With the rapid development of cloud computing technology, the demand for servers is increasing day by day. The long-term uninterrupted operation of a large number of servers will inevitably lead to an increase in failure rate. How to quickly find and deal with server failures has become a technical problem that technicians need to solve.

At present, each server manufacturer sets up a monitoring platform for its own server to detect the failure of its own server, but different types of servers produced by different server manufacturers, and different operating systems installed on the server itself make it necessary to use the corresponding monitoring platform to monitor Fault detection. That is to say, in the existing technology, the functions of the monitoring platforms used for fault detection of different servers are uneven, and most of the monitoring platforms are based on unique protocols, resulting in the inability to perform unified supervision and unified output. When fault detection is performed on servers with various types of operating systems and servers with multiple types of operating systems, the operation process is complicated and the detection efficiency is low.

Therefore, how to improve the convenience and efficiency of server fault detection is a technical problem that those skilled in the art need to solve.

Contents of the invention

In view of this, the object of the present invention is to provide a server fault detection method, which can improve the convenience and efficiency of server fault detection; another object of the present invention is to provide a server fault detection device, equipment and computer Reading storage media all have the above beneficial effects.

In order to solve the above technical problems, the present invention provides a fault detection method for a server, including:

Collect fault data information generated when the server fails;

performing word segmentation processing on each of the fault data information to obtain corresponding keywords;

Determine the matching times of each of the keywords and the preset keywords in the preset word segmentation class library;

The fault value of the server is calculated according to the matched keywords and the corresponding matching times, and the fault condition of the server is determined according to the fault value.

Preferably, the fault data information includes in-band fault data information and out-of-band fault data information; correspondingly, the process of collecting fault data information generated when a server fails specifically includes:

receiving the in-band fault information sent by the operating system of the server and/or a preset monitoring platform;

Obtaining the in-band fault information by running a preset collection script in the server;

receiving the out-of-band fault data information forwarded by the BMC of the server;

Correspondingly, the process of calculating the fault value of the server according to the matched keywords and the corresponding matching times, and determining the fault condition of the server according to the fault value, specifically includes:

Calculate the fault value of the server according to the matched keywords corresponding to the in-band fault data information and the out-of-band fault data information and the corresponding matching times, and determine the server according to the fault value failure conditions.

Preferably, after the collection of fault data information generated when the server fails, it further includes:

judging the data format type of the fault data information;

If the fault data information is in text format, then store the fault data information in a database, and enter the step of performing word segmentation processing on each of the fault data information to obtain corresponding keywords;

If the fault data information is in graphic format, then identify the text in the fault data information, obtain the fault data information in text format, store the fault data information in text format in the database, and Enter the step of performing word segmentation processing on each of the fault data information to obtain corresponding keywords.

Preferably, further comprising:

Preset the fault alarm upper limit and fault alarm lower limit;

Correspondingly, the process of calculating the fault value of the server according to the matched keywords and the corresponding matching times, and determining the fault condition of the server according to the fault value, specifically includes:

calculating the fault value of the server according to the matched keywords and the corresponding matching times;

If the fault value is greater than the fault alarm lower limit and smaller than the fault alarm upper limit, then determine that the server can automatically repair the fault, and start a preset fault repair program;

If the fault value is greater than the fault alarm upper limit value, it is determined that the server is a fault that cannot be automatically repaired, and a corresponding alarm message is issued.

Preferably, the process of performing word segmentation processing on each of the fault data information to obtain corresponding keywords specifically includes:

judging the language type of the fault data information;

If the fault data information is Chinese, then utilize the Chinese word segmentation tool to carry out word segmentation processing on the fault data information to obtain corresponding keywords;

If the fault data information is in English, use an English word segmentation tool to perform word segmentation processing on the fault data information to obtain corresponding keywords.

Preferably, the process of determining the number of matches between each of the keywords and the preset keywords in the preset word segmentation class library specifically includes:

Determining the removed words in the keywords by using the removed word class library in the preset word segmentation class library, and deleting the removed words;

Using the keyword class library in the participle class library to match the remaining keywords, and determine the matching times of the remaining keywords with the preset keywords.

Preferably, after calculating the fault value of the server according to the matched keywords and the corresponding matching times, and determining the fault condition of the server according to the fault value, further comprising:

Send the fault condition to the target terminal device by email and/or short message.

In order to solve the above technical problems, the present invention also provides a server fault detection device, including:

The collection module is used to collect fault data information generated when the server fails;

A word segmentation module, configured to perform word segmentation processing on each of the fault data information to obtain corresponding keywords;

A matching module, configured to determine the number of matches between each of the keywords and the preset keywords in the preset word segmentation class library;

A determining module, configured to calculate a fault value of the server according to the matched keywords and corresponding matching times, and determine a fault condition of the server according to the fault value.

In order to solve the above technical problems, the present invention also provides a fault detection device for a server, including:

memory for storing computer programs;

A processor, configured to implement the steps of any one of the server fault detection methods described above when executing the computer program.

In order to solve the above-mentioned technical problems, the present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, any one of the above-mentioned server fault detection methods is implemented. A step of.

A server fault detection method provided by the present invention includes: collecting fault data information generated when the server fails; performing word segmentation processing on each fault data information to obtain corresponding keywords; determining each keyword and the preset word segmentation category The matching times of the preset keywords in the library; the fault value of the server is calculated according to the matched keywords and the corresponding matching times, and the fault condition of the server is determined according to the fault value. It can be seen that this method is based on NLP to perform word segmentation matching on the collected fault data information, and calculate the fault value according to the word segmentation matching situation, and then determine the fault condition of the server. The process of word segmentation matching and determining the fault condition can be performed by a unified computer program , there is no need to differentiate the corresponding fault detection program or set up the corresponding monitoring platform for servers of different manufacturers or servers of different operating systems, so this method can improve the efficiency of fault detection for servers of different manufacturers or servers of different operating systems Convenience, thereby improving the efficiency of fault detection on the server.

In order to solve the above-mentioned technical problems, the present invention also provides a server fault detection device, equipment and computer-readable storage medium, all of which have the above-mentioned beneficial effects.

Description of drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the following will briefly introduce the accompanying drawings that need to be used in the description of the embodiments or prior art. Obviously, the accompanying drawings in the following description are only For some embodiments of the present invention, those skilled in the art can also obtain other drawings according to the provided drawings without creative work.

FIG. 1 is a flowchart of a server fault detection method provided by an embodiment of the present invention;

FIG. 2 is a flow chart of another server fault detection method provided by an embodiment of the present invention;

FIG. 3 is a schematic diagram of a specific storage directory structure provided by an embodiment of the present invention;

FIG. 4 is a structural diagram of a server fault detection device provided by an embodiment of the present invention;

FIG. 5 is a structural diagram of a server fault detection device provided by an embodiment of the present invention.

Detailed ways

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

The core of the embodiment of the present invention is to provide a server fault detection method, which can improve the convenience and efficiency of server fault detection; another core of the present invention is to provide a server fault detection device, equipment and computer-readable storage Medium, all have above-mentioned beneficial effect.

In order to enable those skilled in the art to better understand the solution of the present invention, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

FIG. 1 is a flowchart of a server fault detection method provided by an embodiment of the present invention. As shown in Figure 1, a server fault detection method includes:

S10: collecting fault data information generated when the server fails;

In this embodiment, it is first necessary to obtain fault data information generated when the server fails, specifically including obtaining corresponding fault data information through fault files such as logs, scripts, and fault screenshots. It should be noted that, in actual operation, there is no limit to the specific method of collecting the fault data information of the server. For example, the fault data information can be obtained by running a preset collection script in the server and automatically sent back, or it can be directly received The fault data information obtained and sent by the operating system of the server or the preset monitoring platform corresponding to the server.

S20: Perform word segmentation processing on each fault data information to obtain corresponding keywords;

S30: Determine the number of matches between each keyword and the preset keyword in the preset word segmentation class library.

In this embodiment, after the fault data information is acquired, the acquired fault data information is parsed based on natural language processing (NLP). Specifically, before performing NLP parsing, it is necessary to build a word segmentation library for NLP. The word segmentation library is mainly used to define professional vocabulary for server failures for word segmentation. The set word segmentation class library needs to include: word name, part of speech, and fault index; among them, the word name is used to distinguish the names of different faults; the part of speech includes nouns, adjectives, etc., which is convenient for word segmentation tools to distinguish parts of speech; the fault index indicates the result of word segmentation The keyword is the influence coefficient of the fault, which is used to calculate the fault value.

It should be noted that in actual operation, the word segmentation class library includes a general thesaurus and a special thesaurus. The general thesaurus is used for unified analysis of all server failures, and the dedicated thesaurus is used for word segmentation for special failures. After building the word segmentation class library, take the server as the unit and read the fault data information collected in advance and stored in the database one by one in chronological order, call the word segmentation tool to segment the fault data information, and obtain the word segmentation results, that is, keywords; then Each keyword is matched with a preset keyword in the word segmentation class library, and the matched keyword and the matching times corresponding to the keyword are recorded.

S40: Calculate the fault value of the server according to the matched keyword and the corresponding matching times, and determine the fault condition of the server according to the fault value.

Specifically, after determining the keywords matching the preset keywords in the participle class library and the corresponding matching times, the server is used as a unit to calculate the corresponding fault value of the server, and the fault condition of the server is determined according to the fault value , that is, to determine whether the server is abnormal. For example, with a preset threshold, when the fault value is greater than the preset threshold, it means that the server has a fault that needs to be repaired; otherwise, it means that the server fault can be ignored.

A server failure detection method provided by an embodiment of the present invention includes: collecting failure data information generated when the server fails; performing word segmentation processing on each failure data information to obtain corresponding keywords; determining the relationship between each keyword and the preset The matching times of the preset keywords in the word segmentation class library; calculate the fault value of the server according to the matched keywords and the corresponding matching times, and determine the fault condition of the server according to the fault value. It can be seen that this method is based on NLP to perform word segmentation matching on the collected fault data information, and calculate the fault value according to the word segmentation matching situation, and then determine the fault condition of the server. The process of word segmentation matching and determining the fault condition can be performed by a unified computer program , there is no need to differentiate the corresponding fault detection program or set up the corresponding monitoring platform for servers of different manufacturers or servers of different operating systems, so this method can improve the efficiency of fault detection for servers of different manufacturers or servers of different operating systems Convenience, thereby improving the efficiency of fault detection on the server.

On the basis of the above embodiments, this embodiment further explains and optimizes the technical solution. Specifically, in this embodiment, the fault data information includes in-band fault data information and out-of-band fault data information; correspondingly, the collected The process of fault data information generated when the server fails, specifically includes:

Receive in-band fault information sent by the server's operating system and/or preset monitoring platform;

Obtain in-band fault information by running preset collection scripts in the server;

Receive the out-of-band fault data information forwarded by the BMC of the server;

Correspondingly, the process of calculating the fault value of the server according to the matched keywords and the corresponding matching times, and determining the fault condition of the server according to the fault value, specifically includes:

The fault value of the server is calculated according to the matching keywords corresponding to the in-band fault data information and the out-of-band fault data information and the corresponding matching times, and the fault condition of the server is determined according to the fault value.

Specifically, in this embodiment, the collected fault data information includes in-band fault data information and out-of-band fault data information; wherein, there are two ways to collect in-band fault information, and the first way is to passively receive the information from the server through the information forwarding server. Fault data information sent by the operating system or the default monitoring platform of the server; Method 2: By intruding into the operating system of the server, running the preset collection script, collecting the in-band fault information of the server through the preset collection script and automatically returning it. The out-of-band fault information collection adopts the passive receiving method, and the fault data information forwarded by the BMC of the server is passively received through the information forwarding server.

That is to say, in this embodiment, the fault data information includes two parts, that is, the in-band fault data information and the out-of-band fault data information; Words and corresponding matching times, as well as keywords and corresponding matching times between out-of-band fault data information and the preset word segmentation class library, when calculating the fault value, it needs to be based on in-band fault data information and out-of-band fault data The matching keywords corresponding to the information and the matching times corresponding to each keyword calculate the fault value of the server, and then determine the fault condition of the server according to the fault value.

It can be seen that in this embodiment, by obtaining in-band fault data information and out-of-band fault data information, and calculating the fault value of the server based on these two fault data information, the accuracy of calculating the fault value of the server can be improved, thereby enabling Improve the accuracy of server failure detection.

On the basis of the above embodiments, this embodiment further explains and optimizes the technical solution. Specifically, after collecting the fault data information generated when the server fails, this embodiment further includes:

Determine the data format type of the fault data information;

If the fault data information is in text format, then store the fault data information in the database, and enter the step of performing word segmentation processing on each fault data information to obtain corresponding keywords;

If the fault data information is in graphic format, identify the text in the fault data information, obtain the fault data information in text format, store the fault data information in text format in the database, and enter word segmentation processing for each fault data information, The step of obtaining the corresponding keyword.

Specifically, in this embodiment, after collecting the fault data information generated when the server fails, first identify the data format type of the collected fault data information. Wherein, the data format types include text format and graphic format, and log or script files are generally in text format, and fault screenshots are generally in graphic format.

When the fault data information is in text format, store the fault data information in text format in the database, and enter the step of word segmentation processing for each fault data information to obtain the corresponding keywords; if the fault data information is in graphic format, then It is necessary to first recognize the text in the fault data information in graphic format through OCR (Optical Character Recognition, Optical Character Recognition) technology, obtain the fault data information in text format, and then store the fault data information in text format in the database and enter A step of performing word segmentation processing on each fault data information to obtain corresponding keywords.

It can be seen that the method of this embodiment can process the fault data information in text format and graphic format, increase the diversity and richness of fault data information, and determine the fault value based on more diverse fault data information for server Fault detection can further improve the accuracy of fault detection.

On the basis of the above embodiments, this embodiment further explains and optimizes the technical solution. Specifically, this embodiment further includes:

Preset the fault alarm upper limit and fault alarm lower limit;

Correspondingly, the process of calculating the fault value of the server according to the matched keywords and the corresponding matching times, and determining the fault condition of the server according to the fault value, specifically includes:

Calculate the failure value of the server according to the matched keywords and the corresponding matching times;

If the fault value is greater than the lower limit of the fault alarm and less than the upper limit of the fault alarm, it is determined that the server can automatically repair the fault, and a preset fault repair program is started;

If the fault value is greater than the upper limit value of the fault alarm, it is determined that the server is a fault that cannot be automatically repaired, and a corresponding alarm message is issued.

Specifically, in this embodiment, the upper limit value of the fault warning and the lower limit value of the fault warning are further preset; For comparison; if the fault value is greater than the lower limit of the fault alarm and smaller than the upper limit of the fault alarm, that is, the fault value exceeds the lower limit of the fault alarm, but has not yet reached the upper limit of the fault alarm, so it means that the current fault of the server can be automatically repaired fault, so start the preset fault repair program to self-repair the server; if the fault value is greater than the upper limit of the fault alarm, it means that the current fault of the server cannot be automatically repaired, and a corresponding alarm message will be issued to notify the person in charge of the server to repair the fault. Carry out manual fault repair; if the fault value is less than the lower limit of the fault alarm, it means that the current fault of the server can be ignored, so continue to collect fault data information when the server is faulty.

It can be seen that in this embodiment, by further using the calculated fault value to compare with the preset fault alarm upper limit value and fault alarm lower limit value, the fault condition of the server can be further analyzed, and when the fault of the server is a fault that can be automatically repaired , start the preset fault repair program to self-repair the server, thus relatively reducing the operations of technicians to perform troubleshooting and repair, and relatively reducing the workload of technicians.

On the basis of the above embodiments, this embodiment further explains and optimizes the technical solution. Specifically, in this embodiment, the process of word segmentation processing is performed on each fault data information to obtain corresponding keywords, which specifically includes:

Determine the language type of the fault data information;

If the fault data information is in Chinese, use the Chinese word segmentation tool to perform word segmentation processing on the fault data information to obtain the corresponding keywords;

If the fault data information is in English, use the English word segmentation tool to perform word segmentation processing on the fault data information to obtain corresponding keywords.

Specifically, in this embodiment, in the process of performing word segmentation processing on each fault data information to obtain corresponding keywords, first analyze and judge the language type of the fault data information, and the language types include Chinese and English, so it is necessary to Use the corresponding word segmentation tool to perform word segmentation processing on the fault data information, and obtain the corresponding keywords. As a preferred embodiment, the Chinese word segmentation tool can be specifically Ansj; the English word segmentation tool can be specifically NLTK; this embodiment does not limit the type of the specific word segmentation tool used.

It can be seen that in this embodiment, by further distinguishing the language types of fault data information, and using word segmentation tools corresponding to the language types to perform word segmentation processing on fault data information, the accuracy of keywords obtained by word segmentation can be further improved, thereby improving the calculation of faults. The accuracy of the value, thereby improving the accuracy of server failure detection.

On the basis of the above-mentioned embodiments, this embodiment further explains and optimizes the technical solution. Specifically, in this embodiment, the matching of each keyword with the preset keywords in the preset participle class library is determined. process, including:

Utilize the removed part of speech library in the pre-set participle class library to determine the removed words in the keywords, and delete the removed words;

Using the keyword class library in the word segmentation class library to match the remaining keywords, and determine the matching times of the remaining keywords with the preset keywords.

Specifically, in this embodiment, when the participle class library is set, the removal of the part of speech library is further set, and what is set in the removal of the part of speech library is a preset removal word, and the preset removal word includes words that are irrelevant to determining whether the server is faulty, And the words that will cause interference/fuzziness to determine whether the server is faulty; then each keyword of the fault data information is matched with each preset removal word in the word class library to determine the removal word in the keyword, and The matched removed words are deleted to obtain the remaining keywords; then the remaining keywords are matched using the keyword class library in the word segmentation class library to determine the remaining keywords and the preset keywords in the keyword class library matching keywords, and determine the corresponding matching times.

That is to say, this embodiment screens out the removed words in the keywords in advance, and then uses the remaining keywords to match with the keyword class library in the word segmentation class library, and then determines the preset keyword according to the remaining keywords. The matching keywords and the corresponding matching times can improve the accuracy of determining the keywords matching the preset keywords in the keyword class library in the participle class library and the corresponding matching times, thereby improving the computing power of the server. The accuracy of the fault value, thereby improving the accuracy of server fault detection.

On the basis of the above embodiments, this embodiment further explains and optimizes the technical solution. Specifically, this embodiment calculates the fault value of the server according to the matched keywords and the corresponding matching times, and calculates the fault value according to the fault value After determining the fault condition of the server, it further includes:

Send the fault condition to the target terminal device by email and/or SMS.

Specifically, in this embodiment, the target terminal device that receives the fault condition is set in advance, and then the fault value of the server is calculated according to the matched keyword and the corresponding matching times, and the fault state of the server is determined according to the fault value, That is, after determining the fault condition of the fault detection of the server, the fault condition is further sent to the target terminal device by mail and/or short message; so that the technician can remotely obtain the fault detection situation of the server through the target terminal device, thereby further Improve the convenience of server failure detection.

In order to enable those skilled in the art to better understand the technical solutions in the present application, the technical solutions in the embodiments of the present application will be described in detail below in combination with actual application scenarios. Figure 2 is a flow chart of another server fault detection method provided by the embodiment of the present invention; The specific process of the fault detection method of the server is as follows:

1. Preparation stage

Enter the device information of the client server that needs to perform fault detection, including server name, server address (IP), fault alarm upper limit and fault alarm lower limit, into the server information table. The table structure is as follows:

Table 1 server information table

serial number name of server IP Fault alarm upper limit Fault alarm lower limit Current alarm value

Create NLP word classification library, including the creation of general thesaurus and exclusive thesaurus. Both the general thesaurus and the exclusive thesaurus are set with the keyword class library and the word class library is removed. Preset keywords are added to the keyword class library and the word class library is removed. Add preset removal words in the general lexicon, store common fault identification keywords in the general lexicon, and add fault identification keywords that cannot be applied to the general lexicon in the exclusive lexicon. When entering each keyword, it is necessary to add the part of speech (such as noun, adjective etc.) and failure index. Taking Ansj Chinese word segmentation as an example, manual.dic and special.dic need to be created for the general thesaurus and special thesaurus, and the general and special thesaurus are loaded into the configuration file, while the fault index corresponding to the keyword needs to be stored separately into the database table.

2. Information collection

The collected fault data information includes in-band fault data information and out-of-band fault data information. Specifically, the target address of the pre-configured information forwarding server is the server IP where the information collection module is located, that is, the IP of the target server; the information collection program is started, and the information collection script is automatically and regularly sent to each client server, while maintaining the connection with the information forwarding server The link is unblocked, waiting to receive fault data information; after the client server receives the command to run the information collection script, it will automatically run the information collection script and send the execution result back to the target server; The operating system or the corresponding preset monitoring platform obtains the corresponding fault data information, and forwards it to the target server through the information forwarding server. When the BMC of the client server detects the fault data information, it forwards the fault data information to the target server through the information forwarding server.

As shown in Figure 3, it is a schematic diagram of a specific storage directory structure provided by the embodiment of the present invention. The collected fault data information is stored in the form of a file, and the in-band fault data information is stored in the in_band folder. The data information is stored in the out_of_band folder, and the fault data information regardless of in-band or out-of-band is stored in units of servers (for example: server01), and the failure of the same server is stored in units of timestamps (for example: 1594532194).

3. Data storage

Start the data storage program and enter the storage directory of fault data information.

Scan the in_band folder and the out_of_band folder respectively, obtain the list of all folders under the current directory of the in_band folder and out_of_band folder respectively, start from the first folder, enter the server01 directory, scan the time stamp folder under the directory, and read Take the files in each directory to get the fault data information and identify the data format type; if it is in text format, write the fault data information directly into the server fault information table; if it is in graphic format, call OCR analysis to identify The text in the fault data information in graphic format is obtained to obtain the fault data information in text format and stored in the server fault information table. It should be noted that when the fault data information is stored in the server fault information table, corresponding storage needs to be performed according to the in-band fault data information and the out-of-band fault data information.

Table 2 Servers

serial number name of server In-band fault information out-of-band fault information update time

4. NLP analysis

Start the NLP analysis, read the contents of the server fault information table one by one, and call the Chinese word segmentation tool (example: Ansj) or English word segmentation tool (example: NLTK) to analyze the in-band fault data according to the language type of the read fault data information. Word segmentation is performed on the information and out-of-band fault data information respectively, and the word segmentation result information (keywords) is saved in the fault information word segmentation result table. The table structure looks like this:

Table 3 Fault information word segmentation result table

serial number name of server In-band word segmentation results Out-of-band segmentation result update time

Read the fault information word segmentation result table, and match the contents of the two columns of in-band word segmentation result and out-of-band word segmentation result with the preset word segmentation class library. Specifically, it includes using the general thesaurus and the exclusive thesaurus for matching, and removing the removed words in the keywords, saving the matched keywords and the number of matches corresponding to each keyword, and storing them in the fault matching result table. The table structure looks like this:

Table 4 Fault matching result table

serial number name of server In-band keyword matching results Out-of-band keyword matching results update time

5. Fault calculation

Start the fault calculation program, the fault calculation program takes the client server as the unit, reads the fault matching result table, and calculates the fault value of the server. In this embodiment, the method of calculating the fault value of the server is as follows:

Among them, f _m represents the fault value of the server, a represents the number of keywords matching the preset keywords in the word segmentation class library in the in-band fault data information, and b represents the number of keywords in the out-of-band fault data information and the word segmentation class library The number of keywords matched by preset keywords, c (c=a+b) represents the total number of keywords matched with the preset keywords in the word segmentation class library in the in-band and out-of-band fault data information; δ represents the in-band The coefficient level of the fault; γ represents the coefficient level of the out-of-band fault; n _i represents the matching number of keyword i in a matching keywords; n _j represents the matching number of keyword j in b matching keywords x _i represents the matching coefficient of keyword i in a matching keywords; x _j represents the matching coefficient of keyword j in b matching keywords; x _k (x _k = x _i ∪ x _j ) Indicates the matching coefficient of keyword k (k=i∪j) among the c matching keywords.

Store the calculated current failure value of the client server into the server information table (Table 1).

6. Results push

Start the result push program, read the fault alarm upper limit value f _max (m) and the fault alarm lower limit value f _min (m) of each client server entered in the server information table; read the current fault value f of the client server _m , judge whether it is satisfied (f _max (m)＞f _m )&&(f _m ＞f _min (m)); if it is satisfied, it is determined that the client server can automatically repair the fault, so the preset fault repair program is started automatically Repair the failure; if f _max (m)<f _m is judged to be satisfied, it means that the current failure of the server cannot be automatically repaired. Therefore, an alarm message is sent to the target terminal device by email and/or short message, and the person in charge of the server is notified to repair the failure of the server. Carry out manual fault repair; if none of the above conditions are satisfied, that is, f _min (m)>f _m , continue to obtain fault data information of the client server.

A server fault detection method provided by an embodiment of the present invention is based on word segmentation matching of collected fault data information based on NLP, and calculates the fault value according to the word segmentation matching situation, and then determines the fault condition of the server, word segmentation matching and determining the fault The process of the situation can be performed using a unified computer program, and there is no need to differentiate the corresponding fault detection program or set the corresponding monitoring platform for servers of different manufacturers or servers of different operating systems, so this method can improve the detection of servers of different manufacturers Or the convenience of fault detection for servers with different operating systems, thereby improving the efficiency of fault detection for servers.

The embodiment of a server fault detection method provided by the present invention has been described in detail above, and the present invention also provides a server fault detection device, equipment, and computer-readable storage medium corresponding to the method, because The embodiments of the device, equipment, and computer-readable storage medium correspond to the embodiments of the method, so for the embodiments of the device, equipment, and computer-readable storage medium, please refer to the description of the embodiments of the method, and will not be repeated here. .

FIG. 4 is a structural diagram of a server fault detection device provided by an embodiment of the present invention. As shown in FIG. 4 , a server fault detection device includes:

A collection module 41, configured to collect fault data information generated when the server fails;

The word segmentation module 42 is used to carry out word segmentation processing to each fault data information to obtain corresponding keywords;

Matching module 43, is used for determining the matching times of each keyword and the preset keyword in the pre-set participle class library;

The determining module 44 is configured to calculate the fault value of the server according to the matched keyword and the corresponding matching times, and determine the fault condition of the server according to the fault value.

The server fault detection device provided by the embodiment of the present invention has the beneficial effect of the above server fault detection method.

As a preferred embodiment, the acquisition module specifically includes:

The first collection sub-module is used to receive the in-band fault information sent by the operating system of the server and/or the preset monitoring platform;

The second collection sub-module is used to obtain in-band fault information by running a preset collection script in the server;

The third acquisition sub-module is used to receive the out-of-band fault data information forwarded by the BMC of the server;

Correspondingly, the matching module specifically includes:

The determination sub-module is used to calculate the fault value of the server according to the corresponding matching keywords and the corresponding matching times of the in-band fault data information and the out-of-band fault data information, and determine the fault condition of the server according to the fault value.

As a preferred embodiment, a server fault detection device further includes:

Judging module, used for judging the data format type of fault data information;

The first execution module is used to store the fault data information in the database and call the word segmentation module if the fault data information is in text format;

The second execution module is used to identify the text in the fault data information if the fault data information is in graphic format, obtain the fault data information in text format, store the fault data information in text format in the database, and call the word segmentation module .

As a preferred embodiment, a server fault detection device further includes:

The alarm value setting module is used to pre-set the upper limit value of the fault alarm and the lower limit value of the fault alarm;

Correspondingly, determine the modules, including:

The calculation sub-module is used to calculate the failure value of the server according to the matched keywords and the corresponding matching times;

The third execution module is used to determine that the server can automatically repair the fault if the fault value is greater than the fault alarm lower limit value and smaller than the fault alarm upper limit value, and start a preset fault repair program;

The fourth execution module is used to determine that the server is a failure that cannot be automatically repaired if the failure value is greater than the upper limit of the failure alarm, and send corresponding alarm information.

As a preferred embodiment, a server fault detection device further includes:

The sending module is used to send the fault condition to the target terminal device by email and/or short message.

FIG. 5 is a structural diagram of a server fault detection device provided by an embodiment of the present invention. As shown in FIG. 5 , a server fault detection device includes:

memory 51 for storing computer programs;

The processor 52 is configured to implement the steps of the above-mentioned server fault detection method when executing the computer program.

The server fault detection device provided by the embodiment of the present invention has the beneficial effect of the above server fault detection method.

In order to solve the above technical problems, the present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above-mentioned server fault detection method are realized.

The computer-readable storage medium provided by the embodiment of the present invention has the beneficial effect of the above-mentioned server fault detection method.

The server fault detection method, device, equipment and computer-readable storage medium provided by the present invention have been introduced in detail above. In this paper, specific examples are used to illustrate the principles and implementation modes of the present invention, and the descriptions of the above examples are only used to help understand the methods and core ideas of the present invention. It should be pointed out that for those skilled in the art, without departing from the principle of the present invention, some improvements and modifications can be made to the present invention, and these improvements and modifications also fall within the protection scope of the claims of the present invention.

Each embodiment in the description is described in a progressive manner, each embodiment focuses on the difference from other embodiments, and the same and similar parts of each embodiment can be referred to each other. As for the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and for the related information, please refer to the description of the method part.

Professionals can further realize that the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, computer software or a combination of the two. In order to clearly illustrate the possible For interchangeability, in the above description, the composition and steps of each example have been generally described according to their functions. Whether these functions are executed by hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art may use different methods to implement the described functions for each specific application, but such implementation should not be regarded as exceeding the scope of the present invention.

Claims (10)

1. A method for detecting a failure of a server, comprising:

collecting fault data information generated when a server fails;

performing word segmentation processing on each fault data information to obtain corresponding keywords;

determining the matching times of each keyword and a preset keyword in a preset word classification library;

calculating a fault value of the server according to the matched keywords and the corresponding matching times, and determining the fault condition of the server according to the fault value;

the fault value of the server is calculated according to the matched keywords and the corresponding matching times, and is specifically calculated according to the following formula:

wherein, f _m For the fault value, a is the number of the keywords matched with the preset keywords in the word class library in the in-band fault data information, b is the number of the keywords matched with the preset keywords in the word class library in the out-of-band fault data information, c = a + b, and c is the number of the keywords matched with the preset keywords in the word class library in the in-band and out-of-band fault data informationThe total number of the keywords matched with the preset keywords in the word class library is delta, gamma is the coefficient grade of the in-band fault, gamma is the coefficient grade of the out-of-band fault, and n is _i The matching number n of the keywords i in the keywords which are matched for a _j The matching number, x, of the keywords j in the keywords that are b matched _i Matching coefficient, x, of keyword i of the keywords matched for a _j A matching coefficient, x, of a keyword j of the b matched keywords _k ＝x _i ∪x _j ，x _k And matching coefficients of key words k in the key words matched with the c key words, wherein k = i ≧ j.

2. The method of claim 1, wherein the fault data information comprises the in-band fault data information and the out-of-band fault data information; correspondingly, the process of collecting fault data information generated when the server fails specifically includes:

receiving the in-band fault information sent by an operating system and/or a preset monitoring platform of the server;

acquiring the in-band fault information by running a preset acquisition script in the server;

receiving the out-of-band fault data information forwarded by the BMC of the server;

correspondingly, the process of calculating the fault value of the server according to the matched keyword and the corresponding matching times and determining the fault condition of the server according to the fault value specifically includes:

and calculating a fault value of the server according to the matched keywords and the corresponding matching times which respectively correspond to the in-band fault data information and the out-of-band fault data information, and determining the fault condition of the server according to the fault value.

3. The method of claim 1, wherein after collecting failure data information generated when the server fails, further comprising:

judging the data format type of the fault data information;

if the fault data information is in a text format, storing the fault data information into a database, and performing word segmentation processing on each fault data information to obtain a corresponding keyword;

if the fault data information is in a graphic format, identifying characters in the fault data information to obtain the fault data information in a text format, storing the fault data information in the text format into the database, and performing word segmentation processing on each fault data information to obtain corresponding keywords.

4. The method of claim 1, further comprising:

presetting a fault alarm upper limit value and a fault alarm lower limit value;

calculating the fault value of the server according to the matched keywords and the corresponding matching times;

if the fault value is larger than the lower fault alarm limit value and smaller than the upper fault alarm limit value, determining that the server can automatically repair the fault, and starting a preset fault repair program;

if the fault value is larger than the fault alarm upper limit value, determining that the server cannot automatically repair the fault, and sending corresponding alarm information.

5. The method according to claim 1, wherein the process of performing word segmentation processing on each fault data message to obtain a corresponding keyword specifically includes:

judging the language type of the fault data information;

if the fault data information is Chinese, performing word segmentation processing on the fault data information by using a Chinese word segmentation tool to obtain a corresponding keyword;

and if the fault data information is English, performing word segmentation processing on the fault data information by using an English word segmentation tool to obtain a corresponding keyword.

6. The method according to claim 1, wherein the process of determining the matching times of each keyword with preset keywords in a preset segmentation class library specifically comprises:

determining a removed word in the keyword by using a removed word library in the preset word class library, and deleting the removed word;

and matching the rest keywords by using the keyword class library in the word segmentation class library to determine the matching times of the rest keywords and the preset keywords.

7. The method according to any one of claims 1 to 6, wherein after the calculating a failure value of the server according to the matched keyword and the corresponding matching times and determining a failure condition of the server according to the failure value, further comprising:

and sending the fault condition to the target terminal equipment in a mail and/or short message mode.

8. A failure detection apparatus for a server, comprising:

the acquisition module is used for acquiring fault data information generated when the server fails;

the word segmentation module is used for carrying out word segmentation processing on each fault data information to obtain corresponding keywords;

the matching module is used for determining the matching times of each keyword and preset keywords in a preset segmentation class library;

the determining module is used for calculating a fault value of the server according to the matched keywords and the corresponding matching times and determining the fault condition of the server according to the fault value;

the fault value of the server is calculated according to the matched keywords and the corresponding matching times, and is specifically calculated according to the following formula:

wherein f is _m For the fault value, a is the number of the keywords matched with the preset keywords in the segmentation class library in the in-band fault data information, b is the number of the keywords matched with the preset keywords in the segmentation class library in the out-of-band fault data information, c = a + b, c is the total number of the keywords matched with the preset keywords in the segmentation class library in the in-band and out-of-band fault data information, delta is the coefficient grade of the in-band fault, gamma is the coefficient grade of the out-of-band fault, n is the coefficient grade of the out-of-band fault _i The matching number n of the keywords i in the keywords which are matched for a _j The matching number, x, of the keywords j in the keywords that are b matched _i Matching coefficient, x, of keyword i of the keywords matched for a _j Matching coefficient, x, of keyword j in the keywords matched for b _k ＝x _i ∪x _j ，x _k And k = i ≡ j is a matching coefficient of a keyword k in the c matched keywords.

9. A failure detection apparatus of a server, characterized by comprising:

a memory for storing a computer program;

processor for implementing the steps of the method of fault detection of a server according to any of claims 1 to 7 when executing said computer program.

10. A computer-readable storage medium, characterized in that a computer program is stored thereon, which computer program, when being executed by a processor, carries out the steps of a method of fault detection of a server according to any one of claims 1 to 7.

Images