Note: Descriptions are shown in the official language in which they were submitted.
METHOD AND SYSTEM FOR AUTOMATICALLY MONITORING BUSINESS
SYSTEMS
BACKGROUND OF THE INVENTION
Technical Field
[0001] The present invention relates to the technical field of E-business, and
more particularly
to a method and a system for automatically monitoring business systems.
Description of Related Art
[0002] In early days where systems are of relatively small scales, maintenance
and operation of
these systems mainly relied on manual processing of maintenance and operation
staff. With the
rapid growth of business and increasing diverseness of services, building
networks has become
more and more complicated. A single system may involve multiple devices and
need to deploy
plural instances, making manual checking and locating problems by maintenance
and operation
staff become more difficult and more inefficient. Increase of the number of
devices also means
that communication between maintenance and operation teams are becoming
costly. Since
various logs and alerts are distributed over different devices, checking
statistical data and then
prompting relevant businesses to take action are time- and effort-consuming.
[0003] The traditional business systems lack for a consistent monitoring
solution, and usually
rely on third-party systems for measurement of various indicators, making it
complicated for
businesses to adopt. Besides, since its indicator monitoring data and alert-
related configuration
are scattered over different systems, though such known systems can configure
several
monitoring alerts, it is difficult for them to see the root causes of problems
and to estimate
affected ranges. Additionally, even if such known systems give alerts, the
alert information is
nevertheless too blurred for maintenance and operation staff to locate
problems directly. In this
case, businesses have to check all the suspected part one by one, making
troubleshooting time-
and effort-consuming.
[0004] Hence, there is a pressing need for a method or a system that is easy
to adopt and can
provide business systems with automated inspection to fast identify problems.
1
Date recue / Date received 2021-12-09
SUMMARY OF THE INVENTION
[0005] To address the defects of the prior art, the objective of the present
invention is to provide
a method and a system for automatically monitoring business systems that solve
the technical
problems raised from complexity and ineffectiveness in recognizing and
locating abnormalities
of business services as seen in existing methods for monitoring business
systems.
[0006] In a first aspect, the present invention provides a method for
automatically monitoring
business systems. The method comprises: automatically collecting dependency
property
information of a business system through JAR packages of a monitoring and
inspecting
component and reporting the dependency property information to a service end;
[0007] after the service end configures the relevant business system according
to the
dependency property information, automatically generating inspection case
tasks and issuing the
inspection case tasks to the monitoring and inspecting component;
[0008] executing the inspection case tasks and reporting inspection results by
the monitoring
and inspecting component; and
[0009] locating problems and giving alerts by the service end according to the
inspection results.
[0010] Further, the configuration comprises: custom configuration of business
system cases,
configuration of user access authority, configuration of sensitivity of early
warning and alerts,
and configuration of inspection task strategies.
[0011] Further, the monitoring and inspecting component according to different
business
system services, configures different parameters and task strategies and
performs task
scheduling on the monitoring and inspecting component, the monitoring and
inspecting
component periodically executes inspection case tasks.
[0012] Further, the custom configuration of business system cases comprises
mapping rules for
inspection tasks, and service names of service machines are acquired through
the mapping rules,
thereby locating inspected abnormal services.
[0013] Further, the service end actively gives troubleshooting instructions to
business nodes
determined as having failures, so as to acquire information required by
relevant troubleshooting,
2
Date recue / Date received 2021-12-09
which includes acquiring memory dump documents and acquiring JVM thread
running states.
[0014] Further, the service end provides component expansion service ports.
According to
different business system services, the service end provides parameters to
configure the
component expansion service ports and control a monitoring and inspecting
component to
monitor different case tasks.
[0015] Further, for indicator-type inspection results, calculation is made
based on baseline
values from historical inspections and the configured sensitivity so as to
determine whether an
early warning and an alert are to be triggered.
[0016] Further, and for middleware-type inspection results, fault sources are
identified by
means of multi-machine cross inspection.
[0017] Further, the monitoring and inspecting component further verifies a
post-alert processed
business problem, and, after verification, gives a feedback instruction to the
service end, to make
the service end cancel the early warning or the alert for the business
problem.
[0018] In another aspect, the present invention provides a system for
automatically monitoring
business systems. The system comprises:
[0019] a monitoring and inspecting component, automatically collecting
dependency property
information of a business system through component JAR packages and reporting
the
dependency property information to a configuration service module, executing
inspection case
tasks issued by the configuration service module, and reporting inspection
results; and
[0020] the configuration service module, configuring relevant business systems
according to
the dependency property information, generating the inspection case tasks
issued to the
monitoring and inspecting component, and receiving the inspection results
reported by the
monitoring and inspecting component so as to locate problems and give an
alert.
[0021] Further, the system further comprises: inspection query module, for
making queries for
inspection case information according to a query interface provided by the
configuration service
module.
[0022] As compared to the prior art, the method and system for automatically
monitoring
business systems of the present invention accomplish the following technical
effects:
[0023] Adoption is easy and less intrusive to existing business systems, and
only needs
3
Date recue / Date received 2021-12-09
introducing JAR packages of the component, by which information related to
service-
dependency link properties can be automatically collected and reported to the
service end
through the JAR packages of the introduced component.
[0024] 2. The present invention supports custom inspection tasks, and is
capable of issuing
operational states of the process nodes processed by the business system
itself in a quasi-real-
time manner, and accurately identifies which process node has failure and
underlying reason
according to predetermined rules.
[0025] 3. By using system sensitivity parameters as the benchmark values for
early warning
and alerts, the present invention eliminates the need of separate
configuration of thresholds for
early warning and alerts, while enhancing accuracy of early warning and
alerts, thereby reducing
the risk of false positives.
BRIEF DESCRIPTION OF THE DRAWINGS
[0026] FIG. 1 is a flowchart of a method for automatically monitoring business
systems
according to one embodiment of the present invention.
[0027] FIG. 2 is a structural diagram of a system for automatically monitoring
business systems
according to one embodiment of the present invention.
DETAILED DESCRIPTION OF THE INVENTION
[0028] The following description details some embodiments of the present
invention, and
illustrations of these embodiments are shown in the accompanying drawings.
Therein, the
embodiments described with reference to the drawings are exemplary and
intended to explain
the present invention, and shall by no means be understood as limitation to
the present invention.
[0029] FIG. 1 is a flowchart of a method for automatically monitoring business
systems
according to one embodiment of the present invention.
[0030] As shown in FIG. 1, a method for automatically monitoring business
systems comprises
the following steps.
[0031] The step Si 1 involves automatically collecting dependency property
information of a
4
Date recue / Date received 2021-12-09
business system through JAR packages of a monitoring and inspecting component
and reporting
the dependency property information to a service end.
[0032] In a business system, the monitoring and inspecting component are
introduced for
collecting dependency service link property information through JAR packages,
and reporting
the dependency property information to a service end. The existing solutions
for business
monitoring, system monitoring, and custom monitoring have to connect multiple
client ends,
and
have to be upgraded separately with the upgrading of these client ends, so the
operation is
relatively complicated. The present invention uses a single component that can
be easily
introduced into business systems in a consistent, less-intrusive manner,
thereby eliminating the
need of additional configuration of information related to service-dependent
link properties of
system at the business system. Instead, adoption can be simply accomplished
using JAR
packages of the monitoring and inspecting component. The JAR packages,
designed as plug-in
elements, support not only business monitoring but also support system
monitoring and custom
expansion monitoring. Therein, the dependency property information
specifically refers to two
things, namely dependency services and property information. The dependency
services mainly
refer to information of data exchange between the present system and the
exterior, including
databases, caches, and external system ports accessed. The property
information refers to
information of addresses or authorization required by linking to databases,
such as domain
names, IP addresses, ports, instance names, and URLs of services. These
details will be sent to
the service end for storage. The monitoring and inspecting component, based on
agreed rules,
automatically detects systems, services, and middleware, including Java
virtual machines, and
Docker containers. For example, it recognizes whether MySQL, Redis, or MQ is
used, initiates
linking according to its property information, and regularly performs
instructions about default
detection.
[0033] At the step S12, the method involves, after the service end configures
the relevant
business system according to the dependency property information,
automatically generating
inspection case tasks and issuing the inspection case tasks to the monitoring
and inspecting
component.
[0034] The monitoring component sends the dependency property information of
individual
Date recue / Date received 2021-12-09
business systems collected in a real-time manner to the service end. After
receiving the
dependency property information from the monitoring and inspecting component,
the service
end first verifies consistency. The latest information received by the service
end is compared
with the previously stored information, and if the information is about a
newly added
dependency service link, updating is performed at the service end, such as
updating the IP
address of a database. After verification, the relevant business system is
configured. With the
received service link property information of an introduced system, the
service end can learn
what are the services the system depends on, and generate corresponding
inspection cases
according to these services.
[0035] The configuration as described above comprises: custom configuration of
business
system cases, configuration of sensitivity of early warning and alerts, access
authority
configuration of the systems, and inspection strategies of the monitoring and
inspecting
component, etc. Therein, different business service ports are configured with
different contents.
[0036] The custom configuration is mainly about configuring services, ports,
or local business
code logics designated to be inspected by the relevant business system, so as
to achieve custom
inspection. Therein, the custom configuration further covers mapping rules of
different business
services, separation into standalone machines according to rules, and their
service names.
Depending on business service ports, different parameters can be configured.
For example, the
parameters required by a money-transfer port include two test bank account
numbers and the
amount to be transferred, etc. The parameters required by an order query
service port include a
user ID, and start and end time. Through these configuration operations, the
subsequent detection
queries and information feedback for the monitoring and inspecting component
are facilitated.
The sensitivity configuration for early warning and alerts is about setting
different sensitivity for
different business systems according to practical needs, with the aim to
reduce false positives in
subsequent determination, and particularly effective to determination for
indicator-type results.
The access authority configuration for the system is about setting access
authorities of
maintenance and operation staff. The inspection strategies provide the
combinations of
inspection items of different business types and the inspection frequencies
for the monitoring
and inspecting component to perform inspection. For example, in order to check
whether a
6
Date recue / Date received 2021-12-09
certain system is normal, inspection for all items may be set as being
performed automatically
at a set interval, and cross inspection may be also set.
[0037] After custom configuration of inspection tasks of business systems, the
disclosed system
can automatically generate inspection case tasks based on the set inspection
rules and task
strategies, and issue the inspection case tasks to the monitoring and
inspecting component.
[0038] Therein, the set rules can be briefed as below:
1) The Jar packages of the monitoring and inspecting components automatically
acquire JNDI name of a database, and automatically partition and extract a
read
JNDI and a write JNDI according to the abbreviation of the relevant system and
JNDI naming rules, automatically. For example, assuming that the system
abbreviation is pgmcms, the corresponding reading and writing JNDI are:
pgmcmsRDS and pgmcmsWDS, respectively. The contents extracted from the
petitioning rules are: pgmcmsRDS 1, pgmcmsRDS 2, .........................
and pgmcmsWDS 1,
pgmcmsWDS 2, ..............................................................
After the data source is determined, the dependent default
verification SQL is performed according to the database type. For example,
SELECT VERSION() is for MySQL, and SELECT*FROM DUAL is for ORACLE.
2) The Jar packages of the monitoring and inspecting components automatically
acquire the redis sharding name, and configure familiar information such as
links in
documents and keys according to the abbreviation of the relevant system and
Redis redis.conf. For example, assuming that the system abbreviation is
pgmcms,
the corresponding redis sharding is pgmcms 1, pgmcms 2, .....
[0039] By setting rule examples, accurate problem locating can be realized.
Assuming that 5
databases (each machine is installed with a database) provide a VIP outgoing
service, it is
impossible to tell the database having problems is on which machine by
directly inspecting VIP.
Therefore, the names of service provided on the individual machines have to be
found according
to the VIP service names. This process is achieved using the mapping rules.
Without rule
mapping, inspection can only depend on domain names known to the public. In
this case, even
if it is certain that there are problems, it is impossible to tell the faulty
machine from all the
machines providing services to the site having that domain name.
7
Date recue / Date received 2021-12-09
[0040] The set task strategies can be briefed as below:
[0041] The monitoring and inspecting component automatically generates a time
interval of
inspection according to the level of the relevant business system (e.g., Level
1, 2, or 3, in which
Level 1 has the top priority, and when a Level 1 system has problems, the
service provision is
interrupted) and whether the relevant system is on a core link. Therein, the
default inspection
interval is the mean of intervals of inspections of the connected systems
having the same system
level, the same link, and the same inspection task dimensions. The inspection
intervals may be
customly configured at the service end.
[0042] In addition, the service end further performs task scheduling on the
monitoring and
inspecting component, so as to control plural monitoring and inspecting
components to conduct
inspections for different task instances. The task scheduling mainly includes:
1. The monitoring
and inspecting component is loaded with all available case tasks of the
introduced system, and
executes them at a set frequency; and 2. the service end may alternatively
initiate a case
inspection on a certain machine in the business system, or it may initiate
case inspections on all
the machines in this system. Since the service end further provides component
expansion service
ports allowing introduce of different business systems and provision of
services of various types,
the only thing has to be done at this time is to issue new parameters through
the ports or adjust
parameters and reschedule tasks, so that operations of different business
services can be easily
recognized.
[0043] In the step S13, the monitoring and inspecting component executes an
inspection task
and reports inspection result.
[0044] The monitoring and inspecting component executes an inspection case and
reports the
results of its execution of this case through RESTFUL. RESTFUL is connected to
different
services through ports defined in advance, and thus features excellent
capability of expansion.
When a depend service inspection reveals abnormality, it is natural to
accurately tell which is
the problematic dependency service. For example, when Redis has a fault, the
inspecting
component will inform the service end of which machine has problems about
Redis services or
is unable to connect link. When the inspecting component finds early warning
about abnormality
of a JVM (java virtual machine), such as early warning for memory usage, the
inspection
8
Date recue / Date received 2021-12-09
component can automatically generate a dump document for memory usage, thereby
facilitating
fasting locating the problems. Objects of inspections may be usability
inspection of an entire
cluster, or may be services of service of every machine in a cluster.
[0045] At the step S14, the service end according to inspection results
locates the problem and
give alerts.
[0046] The service end according to the cases reported in a real-time manner
executes and store
log information, and gives alerts. Specifically, the service end analyzes the
log of inspection
results, and calculates early warning and alerts according to sensitivity
settings of individual
business systems. The results are them sent to the user. Besides, the service
end can further
accurately locate problems. With the previously set rules, the specific
machine providing some
certain service can be identified through mapping, so that it is possible to
accurately locate which
machine is the one having failure.
[0047] Another example is about middleware-type inspection results. Herein,
multi-machine
cross inspection is performed on middleware to identify the final fault
source. For example,
when three JBOSS machines are separately inspected for database usability, if
two of them are
normal and the other one is abnormal, it can be certain that the abnormal
JBOSS machine has
problems.
[0048] A further example is about indicator-type inspection results. To
inspect indicators such
as memory usage or threads, the baseline values based on historical
inspections and the
sensitivity level set at the system are combined to get criteria for giving
early warning and alerts.
For example, assuming that the baseline value of memory usage is 60%, and the
sensitivity level
set for the system is 0.3, early warning will be issued when the memory usage
reaches 78%.
Different systems may have different sensitivity levels to errors. By
introducing system
sensitivity parameters as basic values for early warning and alerts, the need
of directly
configuring thresholds for early warning and alerts is eliminated, and the
accuracy of early
warning and alerts is improved, with reduced false positives.
[0049] Referring to FIG. 2, another embodiment of the present invention
provides a system for
automatically monitoring business systems. The system comprises:
[0050] a monitoring and inspecting component, automatically collecting
dependency property
9
Date recue / Date received 2021-12-09
information of a business system through component JAR packages and reporting
the
dependency property information to a configuration service module, executing
inspection case
tasks issued by the configuration service module, and reporting inspection
results; and
[0051] the configuration service module, configuring relevant business systems
according to
the dependency property information, generating the inspection case tasks
issued to the
monitoring and inspecting component, and receiving the inspection results
reported by the
monitoring and inspecting component so as to locate problems and give an
alert. In addition, the
configuration service module further stores configuration information and
inspection result
records of the individual business systems.
[0052] There is also an inspection query module for making queries for
inspection case
information according to a query interface provided by the configuration
service module. The
queries may include queries for alert notification information, inspection
records, and analysis
records.
[0053] The disclosed system may be introduced through simulating a monitoring
and inspecting
component in a business system, while expanding custom inspection case
configuration and
system sensitivity setting. Then the information of results of the inspection
cases of the business
system can be reported through the monitoring and inspecting component in a
real-time manner,
for the disclosed system to analyze the result information, thereby accurately
detecting whether
any business inspected needs early warning or alert.
[0054] The present invention has been described with reference to the
preferred embodiments
and it is understood that the embodiments are not intended to limit the scope
of the present
invention. Moreover, as the contents disclosed herein should be readily
understood and can be
implemented by a person skilled in the art, all equivalent changes or
modifications which do not
depart from the concept of the present invention should be encompassed by the
appended claims.
Hence, the scope of the present invention shall only be defined by the
appended claims.
Date recue / Date received 2021-12-09
