Shulou(Shulou.com)06/02 Report--

This article will explain in detail how to analyze CMDB software, the content of the article is of high quality, so the editor will share it with you for reference. I hope you will have a certain understanding of the relevant knowledge after reading this article.

The foregoing:

0. Hardware CMDB, that is, asset management for equipment resources, is favored by operators.

Software CMDB, configuration management for business resources, similar to business information management platform, developers, PE and testers are more concerned about here.

1. Hardware CMDB companies are under construction (to varying degrees), software CMDB (business information management platform) systematic management and maintenance is less.

two。 Operation and maintenance staff love hardware CMDB, business information management is only general maintenance, far from the application project-based business maintenance, monitoring efforts are only system and service level).

3. Developers have heavy development tasks and do not have the energy to submit application business information. Everyone is very busy, but what the heck has been developed is often complained by marketing colleagues (real experience).

4. Of course, developers must have requirements analysis and project planning, but maintenance personnel (operators) must understand and maintain relevant business application projects.

5. Communication is very important, operation and maintenance personnel not only passive maintenance, with operational means to treat the project.

Perplexity:

1. When it comes to micro-service development, we need to know where the service is deployed, the current number of instances and resource scheduling, not to mention the log and monitoring data of related applications.

two。 Operation and maintenance personnel carry out alarm inspection or performance analysis or add monitoring, and some services have problems (access error | Service alarm | Application online and offline recovery), but do not know how to find them.

Which relevant developers and testers communicate, especially optimize a sub-project service group to confirm the impact on the fault domain and risk assessment.

3. A large number of applications are online, especially containers and micro-services, resulting in more complex business projects, frequent updates of system service application data strings, and can not be maintained by traditional means.

4. For some explosive (advertising second-casting projects) or flexible applications, the use of application business resources needs to be known by different people.

Recently, the TW business promotion, whether the resources of this project are sufficient for expansion; as a result, the sudden dedicated line is full, affecting other services, which is caused by the promotion business and temporarily increases the bandwidth.

5. Anyway, many of them are spread by word of mouth, only professionals know, the cost of personnel communication is very high, low knowledge will also affect work efficiency. Can you provide a convenient platform for developers,

Operation and maintenance, testers deal with application project business more efficiently.

Application business configuration management-big picture

Application business configuration management system related resources (system layer, host, container, service, service group, application project) association, as well as the corresponding monitoring system relationship.

Some of the results are shown:

1. For business resource configuration management, all belong to 'applications and services'.

two。 Server system management, OS system as the basic resources of applications and services.

3. Based on hardware server maintenance (similar to hardware CMDB), based on server system maintenance, based on application project maintenance, about these three stages. Temporarily arrive between the server system and the application project, the monitoring system is also based on the server system, as the operation and maintenance personnel need this cut-off point more.

4. Docker host node, specially split out.

5. Host node and container instance topology diagram.

Applicable scenarios and environments: (https://github.com/tm-roamer/ctopo/)

(1) scenarios suitable for monitoring the interconnection of ip nodes in a network

(2) Group interconnection access relationship suitable for social networks

6. Container instance.

7. Application project

8. Some service group modules are called by multiple applications. Anyway, it is necessary to quantify differentiated service resources and add service group layers between applications and services.

Micro-service items can also be added here, defined as the application is too large, defined as the service is too small, temporarily here.

9. Application service

Don't write for the time being. Service registration and service discovery, similar to service governance.

CCS is based on consul+register as the service registration discovery and health check center.

10. My little goal.

Only know that these server system resources are normal, the service is also normal, these data are too one-sided, independent can not be associated.

Monitoring requirements (close to the service, alarm merge compression), there is a great need for this service configuration management system.

Alarm merge compression, a project failure, may be multi-directional alarm, need to select a higher-level alarm to report, need to know the relationship chain.

#

How to identify resources:

Hardware-Asset (X86 server, minicomputer, switching routing equipment, firewall, rack, etc.) finally selects SN for unique identification, and device_id is used to store device coding (bar code / QR code label)

Name is the device name (the X86 server is the host name, and the switching device is the device name). The server business IP and management IP are stored in Asset-server, and the switch management IP is stored in Asset-switch.

# hardware devices do not necessarily have an OS system, especially if you purchase Tencent Cloud and Ali Cloud VM hosts, you do not need to provide hardware devices, resulting in Asset-AgentInfo split.

Host information-AgentInfo (kvm, vm, hwvm, X86 server systems) must have OS (linux | windows | unix), and agent is required to install system initialization. End up using hostname

As a unique identity, agent_ip uses the mapping IP (hostname-I). At the beginning of using the IP logo, some systems unexpectedly have four kinds of IP (private network | public network) and have no legal (hostname-I), and finally use hostname.

# if you have encountered a hardware server system where KVM and Docker,kvm are installed at the same time, the host and attached vm are all stored in AgentInfo.

# install agent), but the host system and host node attributes are not consistent, such as DockerNode engine version, network, storage,

# image, it's going to be dismantled anyway. Sometimes when purchasing CCS from a cloud provider, you cannot provide AgentInfo. If you join in maintenance, it is also a split.

# one of the reasons: docking authorization, monitoring and code release applications to facilitate loose docking in the future.

Host node-DockerNode, node_id for the main key (for other purposes), host_on is optional for interfacing with the host system, and finally uses docker_name as the unique identity.

#

Container instance-Container, node_id for the main key (for other purposes); node_on and name end up as unique identifiers.

# the container instance name of different host nodes may be consistent (anyway, this is the case); the life cycle is very short, identifying ID and IP

# the address changes frequently, so temporarily choose node_on and name to identify it.

Resource collection method and change notification.

Server management information collection:

Manual input: fill in the information manually through the web page.

Agent automatic reporting: install Agent agent in the system OS at the same time for data collection and reporting.

Relay automatic reporting: for example, Huawei fushion cute private cloud can read VM-related information through relevant API and store it directly.

Host node and container instance:

Agent automatic report: mainly in the host machine through docker info | docker inspect for data collection and input.

Relay automatic reporting: a third-party platform, through API for information storage.

Service Management:

Manual entry: highly deprecated, for non-standardized and small services.

Agent entry: services must be standardized, services are automatically added during service installation and deployment, generally for traditional services.

Consul service discovery: especially for consul and register services, it is very useful for container services.

Service group management:

Distribution management is generally done through nginx and haproxy, or in other ways.

Haproxy can read service_cluster and related services through the status API page to facilitate automatic storage.

Change Notification:

The above resource changes, it is best to receive the monitoring system, let the relevant personnel understand, liberate our hands.

On how to analyze the CMDB software to share here, I hope that the above content can be of some help to you, can learn more knowledge. If you think the article is good, you can share it for more people to see.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.