Solve the problem of hybrid cloud monitoring and diagnosis difficult to portray the cloud network "gossip map"

In China, after more than ten years of development of cloud computing, enterprises have steadily advanced in IT infrastructure and cloud-native business applications, and the scale of cloud services has increased significantly. However, in the face of complex virtual networks, it is increasingly difficult for enterprises to ensure business security and regulatory requirements. The tools and methods of enterprise monitoring and diagnosis need to be improved to meet the business needs that continue to emerge as the IT infrastructure gradually evolves to the hybrid cloud architecture.
The network challenge of enterprise cloud
In the process of enterprises going to the cloud, the infrastructure is constantly being clouded. With the rapid development of container technology, the application architecture after going to the cloud is gradually becoming micro-service. The high-frequency and dynamic characteristics of cloud services meet the needs of fast-changing services, but more services are going to the cloud, which makes the east-west network traffic surge. The cloudification of infrastructure further blurs the boundaries of the network. Various factors add up to make the business after the cloud. The difficulty of security and operation and maintenance has increased sharply.
The monitoring and diagnosis of the network always exists with the development of the network, and the boundary of the network continues to expand with the cloudification of infrastructure, especially the extension of the network virtualization level. Correspondingly, the tentacles of network monitoring and diagnosis need to increase accordingly. But in a hybrid cloud environment, network monitoring and diagnosis encounter new challenges. The existing monitoring and diagnosis solutions of enterprises cannot cover the ever-increasing east-west traffic, and the resulting network "black box" has become a major obstacle for enterprises to go to the cloud.
Gossip diagram of hybrid cloud network
The ancients used Bagua diagrams to show the dynamic and static states of all natural phenomena. When engaged in production activities, they referred to the hexagrams to seek advantages and avoid disadvantages. As far as the network is concerned, the most important appearance is nothing more than the configuration information and operating status of the network element. To solve the problem of monitoring and diagnosing hybrid cloud networks, it is necessary to observe and inquire about the configuration information and operating status of network elements. A "Gossip Picture" depicting the cloud network is ready to come out.
Points, lines, and areas of the network
In the scenario of network monitoring and diagnosis, if we regard specific data packets as "points", the complete business access path (a stream) is a "line" connected by multiple "points". Streaming information contains key network metadata. However, in an IT environment involving public cloud resources and dedicated line links in multiple data centers and resource pools, the solutions on the market lack a global network status perspective, that is, an associated IaaS The knowledge graph of resources, PaaS resources, and service applications-a panorama of hybrid cloud networks.
DeepFlow of Spruce Networks provides point, line, and surface operation logic for cloud networks, and provides comprehensive information guarantee for monitoring and diagnosis of hybrid clouds. The knowledge graph (network panorama) contains the mapping relationship of the object entities involved in the network, shows a series of different perspectives of the structural topology and the current network traffic, and uses visualization technology and search technology to describe the comprehensive and rich operating information of resource entities. The monitoring and diagnosis of business in the cloud provides guidance based on the diagram.
The drawing of cloud network gossip diagram
To draw a full (eight) scene (hexagram) diagram of a hybrid cloud network, first obtain the topology of the production network, secondly the flow information of the entire network, and finally sort out the relationship between the flow and the network elements according to the mapping relationship between the flow and the network elements. The business-related knowledge graph and identification of abnormal or malicious traffic provide a reliable basis for monitoring and diagnosis after the enterprise goes to the cloud.
The DeepFlow monitoring and diagnosis solution of Spruce Networks is composed of DeepFlow collectors, controllers and data nodes. It meets the network monitoring and diagnosis requirements of various resource pools such as KVM, ESXi, containers, and public clouds; and supports IPv4 and IPv6 protocol environments. Obtaining network streams and data packets in a hybrid cloud environment is not easy. It needs to take into account performance and intrusion. The design of the solution must fully consider the company's existing production environment.
The DeepFlow controller first automatically learns the resources and network topology information in the cloud by connecting with the production environment, especially the cloud platform. By docking key physical devices, DeepFlow completes the first step of drawing a full (eight) scene (hexagram) map.
Considering the scalability of the network monitoring framework, traffic collection and back-end monitoring and diagnosis tools must be decoupled. On the collection side, various types of DeepFlow collectors provide basic information capture capabilities for the entire network traffic collection scheme, supporting physical networks, Resource pool network environments such as KVM, ESXi, containers, and public clouds. For a multi-data center, multi-cloud heterogeneous hybrid cloud infrastructure, the DeepFlow controller implements the management of many collectors under different platforms in a cluster mode. The controller can start and stop mass collectors in seconds, and the collectors can preprocess the flow locally by receiving instructions from the controller. So far, DeepFlow has completed the second step of drawing the whole (eight) scene (gua) map.
As the central brain of the entire system, the controller combines the collected traffic and the already-connected production environment network topology, and uses machine learning and big data technologies to automatically sort out the entire network traffic, combining data types, monitoring indicators, resource attributes, Display method and other dimensions to generate a real network traffic knowledge map of the global business, which is the third step of drawing the DeepFlow full (eight) scene (hexagram) map.
Cloud network gossip diagram display
Users of different roles have their own preferences for the display of panoramic images. Comprehensive coverage of these needs and providing a unified presentation are an important prerequisite for satisfying the interpretation of network hexagrams by all parties.
Enterprises have mastered the network data in the hybrid cloud environment through the DeepFlow platform. The core monitoring indicators are various indicators used to describe network status and performance, mainly including throughput, delay, abnormality, transmission status, etc.; the display methods mainly include distribution, correlation, comparison, and backtracking according to usage scenarios. Network traffic data is also typical time series data, and has corresponding network characteristics. Therefore, DeepFlow provides distributed network time series database services, while satisfying fast writing and data persistence, it continuously optimizes multi-dimensional aggregation query capabilities. Users in different roles on the platform can customize the monitoring panel they care more about and set alarm policies based on the core view.
Deployment of hybrid cloud network monitoring and diagnosis solutions
The overall plan includes three parts: DeepFlow collector, DeepFlow controller, and DeepFlow data node. For the overall plan, it is recommended to plan an independent network monitoring plane for the overall hybrid cloud, and manage the supervision traffic of the hybrid cloud in a unified and independent manner. After the overall plan is completed, the construction can be divided into regions, resource pools, and phases to enable enterprises to have the ability to monitor and diagnose the entire network of hybrid cloud infrastructure to ensure the stable operation of application services.
Since most enterprises already have the ability to monitor traditional physical networks, they usually focus on KVM and container resource pool networks as the first stage of construction, focusing on solving the problem of invisible east-west traffic in the resource pool and realizing the resource pool network Visualization improves the efficiency of O&M troubleshooting and guarantees network service level agreement.
In the second stage, more resource pools are included, deployed in synchronization with the newly expanded resource pool, and connected to the optical and mirrored traffic in the physical network to realize overall data center network monitoring.
The third stage is for public cloud resources in the hybrid cloud. It monitors the network running on it, deploys collectors, has the ability to collect workload or container traffic, and completes the overall monitoring and management of the hybrid cloud IT environment network.
For an already running hybrid cloud environment, it can be deployed and implemented without affecting the operation of the production environment. In network planning, the management, monitoring and distribution plane involved in the DeepFlow platform is reused in the existing network plane. Existing network management plane.
Cloud network gossip chart summary
The DeepFlow hybrid cloud network monitoring and diagnosis solution through effective network traffic collection, data classification and storage, and the close integration of network points, lines, and planes, complements the network for enterprises in the evolution of new IT infrastructure environments such as hybrid cloud and cloud native Monitor the gaps to avoid duplication of construction; respond to the characteristics of cloud native, closely integrate with business, solve actual network monitoring problems, and support enterprise infrastructure to move toward network intelligence.

Gloss PETG Film

Paint surface Gloss PETG Film is a high-quality surface decoration material widely used in the furniture industry. It has the following characteristics:
1. Scratch resistance: The paint surface PETG film adopts special surface treatment technology, making it have high scratch resistance performance. Even in daily use, it can effectively prevent scratches and scratches, and maintain the smoothness and aesthetics of the furniture surface.
2. Oil stain resistance: The paint surface PETG film has excellent oil stain resistance performance. Even if oil spills onto the surface of furniture, it can be easily wiped clean without leaving any marks or stains.
3. Anti fingerprint: The paint surface PETG film adopts special coating technology, which can effectively resist the adhesion of fingerprints. Even if your fingers touch the surface of the furniture, they will not leave obvious fingerprints, maintaining the cleanliness and aesthetics of the furniture.
4. Waterproof: The paint surface PETG film has excellent waterproof performance. Even if soaked or splashed with water, it will not penetrate into the interior of the furniture, protecting its quality and lifespan.
In short, the paint surface PETG film has the characteristics of scratch resistance, oil resistance, fingerprint resistance, and waterproofing, making the furniture surface more durable, easy to clean, and able to maintain long-term aesthetics. It is an ideal surface decoration material for furniture and has been loved and favored by a large number of consumers.

gloss petg film,paint surface furniture film,Gloss film, furniture film, furniture decorative film

Dongguan Sunyo Plastic Co., Ltd , https://www.sunyopet.com