Please open Telegram to view this post
VIEW IN TELEGRAM
800 系列网络适配器（高达 100GbE）
800 系列网络适配器至高可提供 100 GbE 的速度，并可用于 PCIe 和 OCP 外形。800 系列的新功能和改进功能包括 Application Device Queues (ADQ)、Dynamic Device Personalization (DDP) 以及对 iWARP 和 RoCEv2 远程直接内存访问 (RDMA) 的支持，可提供工作负载优化的性能以及满足不断变化的网络要求的灵活性。从 800 系列中进行选择，以处理高性能服务器工作负载，例如 NFV、存储、HPC-AI 和混合云。
800 系列网络适配器至高可提供 100 GbE 的速度，并可用于 PCIe 和 OCP 外形。800 系列的新功能和改进功能包括 Application Device Queues (ADQ)、Dynamic Device Personalization (DDP) 以及对 iWARP 和 RoCEv2 远程直接内存访问 (RDMA) 的支持，可提供工作负载优化的性能以及满足不断变化的网络要求的灵活性。从 800 系列中进行选择，以处理高性能服务器工作负载，例如 NFV、存储、HPC-AI 和混合云。
800 Series Network Adapters (up to 100GbE)
The 800 Series network adapters deliver speeds up to 100 GbE and are available in PCIe and OCP form factors. New and improved features of the 800 Series include Application Device Queues (ADQ), Dynamic Device Personalization (DDP), and support for iWARP and RoCEv2 Remote Direct Memory Access (RDMA), providing workload-optimized performance and the flexibility to meet changing network requirements. Choose from the 800 Series to handle high-performance server workloads such as NFV, storage, HPC-AI, and hybrid cloud.
The 800 Series network adapters deliver speeds up to 100 GbE and are available in PCIe and OCP form factors. New and improved features of the 800 Series include Application Device Queues (ADQ), Dynamic Device Personalization (DDP), and support for iWARP and RoCEv2 Remote Direct Memory Access (RDMA), providing workload-optimized performance and the flexibility to meet changing network requirements. Choose from the 800 Series to handle high-performance server workloads such as NFV, storage, HPC-AI, and hybrid cloud.
Jeddak数据安全沙箱
为了应对上述挑战，火山引擎云安全的安全研究团队推出了Jeddak数据安全沙箱，沙箱以可信执行环境（Trusted Execution Environment，TEE）为核心，为用户提供安全、高效、通用的广告精筛、归因分析等隐私计算能力，现已服务于某头部金融信贷公司，为该客户在国内某广告平台投放过程的用户隐私，提供安全保障。
为了应对上述挑战，火山引擎云安全的安全研究团队推出了Jeddak数据安全沙箱，沙箱以可信执行环境（Trusted Execution Environment，TEE）为核心，为用户提供安全、高效、通用的广告精筛、归因分析等隐私计算能力，现已服务于某头部金融信贷公司，为该客户在国内某广告平台投放过程的用户隐私，提供安全保障。
Jeddak Data Security Sandbox
In order to meet the above challenges, the security research team of Volcano Engine Cloud Security launched the Jeddak Data Security Sandbox. The sandbox is based on the Trusted Execution Environment (TEE) and provides users with safe, efficient and universal privacy computing capabilities such as advertising screening and attribution analysis. It is currently serving a leading financial credit company, providing security protection for the user privacy of the customer during the delivery process of a domestic advertising platform.
In order to meet the above challenges, the security research team of Volcano Engine Cloud Security launched the Jeddak Data Security Sandbox. The sandbox is based on the Trusted Execution Environment (TEE) and provides users with safe, efficient and universal privacy computing capabilities such as advertising screening and attribution analysis. It is currently serving a leading financial credit company, providing security protection for the user privacy of the customer during the delivery process of a domestic advertising platform.
心跳源检测异常
顾名思义，通过心跳源，初步发现异常。通常心跳变化会有三类消息，update消息，delete消息和insert消息。心跳逻辑在于，正常情况下SA服务端与NC建立长连接，每数秒缓存一次心跳，每几分钟打包上报一次，但当NC异常时，长连接感知后，立即上报异常，并修改路由表。所以心跳异常做到秒级感知。
顾名思义，通过心跳源，初步发现异常。通常心跳变化会有三类消息，update消息，delete消息和insert消息。心跳逻辑在于，正常情况下SA服务端与NC建立长连接，每数秒缓存一次心跳，每几分钟打包上报一次，但当NC异常时，长连接感知后，立即上报异常，并修改路由表。所以心跳异常做到秒级感知。
Heartbeat source detection anomaly
As the name implies, the anomaly is initially detected through the heartbeat source. Usually, there are three types of messages for heartbeat changes: update messages, delete messages, and insert messages. The heartbeat logic is that under normal circumstances, the SA server establishes a long connection with the NC, caches the heartbeat every few seconds, and packages and reports it every few minutes. However, when the NC is abnormal, the long connection is detected, the anomaly is immediately reported, and the routing table is modified. Therefore, the heartbeat anomaly can be detected in seconds.
As the name implies, the anomaly is initially detected through the heartbeat source. Usually, there are three types of messages for heartbeat changes: update messages, delete messages, and insert messages. The heartbeat logic is that under normal circumstances, the SA server establishes a long connection with the NC, caches the heartbeat every few seconds, and packages and reports it every few minutes. However, when the NC is abnormal, the long connection is detected, the anomaly is immediately reported, and the routing table is modified. Therefore, the heartbeat anomaly can be detected in seconds.
异常排除
排除非物理机器，将系统中暂时不关注的VM等产生的异常信息排除掉。
排除非业务状态的机器，如装机状态中的，包括生产中，维修中，迁移中，重装中，销毁中，重启中，无管控状态，只监控正常状态的机器。
排除非正在工作的机器，如非working状态机器。异常排除
排除非物理机器，将系统中暂时不关注的VM等产生的异常信息排除掉。
排除非业务状态的机器，如装机状态中的，包括生产中，维修中，迁移中，重装中，销毁中，重启中，无管控状态，只监控正常状态的机器。
排除非正在工作的机器，如非working状态机器。
排除非物理机器，将系统中暂时不关注的VM等产生的异常信息排除掉。
排除非业务状态的机器，如装机状态中的，包括生产中，维修中，迁移中，重装中，销毁中，重启中，无管控状态，只监控正常状态的机器。
排除非正在工作的机器，如非working状态机器。异常排除
排除非物理机器，将系统中暂时不关注的VM等产生的异常信息排除掉。
排除非业务状态的机器，如装机状态中的，包括生产中，维修中，迁移中，重装中，销毁中，重启中，无管控状态，只监控正常状态的机器。
排除非正在工作的机器，如非working状态机器。
Abnormal exclusion
Exclude non-physical machines and exclude abnormal information generated by VMs that are not currently being paid attention to in the system.
Exclude machines in non-business states, such as those in the installation state, including those in production, maintenance, migration, reinstallation, destruction, restart, and no control state, and only monitor machines in normal state.
Exclude machines that are not working, such as machines in non-working state.
Exclude non-physical machines and exclude abnormal information generated by VMs that are not currently being paid attention to in the system.
Exclude machines in non-business states, such as those in the installation state, including those in production, maintenance, migration, reinstallation, destruction, restart, and no control state, and only monitor machines in normal state.
Exclude machines that are not working, such as machines in non-working state.
网络集线器
播报
讨论
上传视频
数据通信系统中的基础设备
网络集线器是数据通信系统中的基础硬件设备，又称多口中继器，属于物理层设备，工作于OSI参考模型第一层。其主要应用于局域网（LAN）环境，通过电器互联构建物理上的星型或树型网络结构，功能类似于多端口信号中继器。
播报
讨论
上传视频
数据通信系统中的基础设备
网络集线器是数据通信系统中的基础硬件设备，又称多口中继器，属于物理层设备，工作于OSI参考模型第一层。其主要应用于局域网（LAN）环境，通过电器互联构建物理上的星型或树型网络结构，功能类似于多端口信号中继器。
Network hub
Broadcast
Discussion
Upload video
Basic equipment in data communication system
Network hub is the basic hardware equipment in data communication system, also known as multi-port repeater, which belongs to physical layer equipment and works at the first layer of OSI reference model. It is mainly used in local area network (LAN) environment, and constructs physical star or tree network structure through electrical interconnection, and its function is similar to multi-port signal repeater.
Broadcast
Discussion
Upload video
Basic equipment in data communication system
Network hub is the basic hardware equipment in data communication system, also known as multi-port repeater, which belongs to physical layer equipment and works at the first layer of OSI reference model. It is mainly used in local area network (LAN) environment, and constructs physical star or tree network structure through electrical interconnection, and its function is similar to multi-port signal repeater.
该设备的主要功能是对接收到的电信号进行再生、整形和放大，以扩展网络传输距离。由于缺乏交换机具备的“智能记忆”能力，其数据传输采用广播方式，将数据包发送至所有相连节点。根据信号处理方式可分为无源、有源和智能三类：无源集线器仅转发信号，有源集线器可放大信号延长传输距离，智能集线器则具备网络管理与路由功能。按结构和功能可分为未管理型、堆叠式及底盘集线器，从局域网应用角度则分为单中继、多网段、端口交换式等五类。其硬件特性决定了其在早期网络架构中的基础连接作用。
The main function of this device is to regenerate, shape and amplify the received electrical signals to extend the network transmission distance. Due to the lack of the "intelligent memory" capability of the switch, its data transmission adopts broadcast mode to send data packets to all connected nodes. According to the signal processing method, it can be divided into three categories: passive, active and intelligent: passive hubs only forward signals, active hubs can amplify signals to extend the transmission distance, and intelligent hubs have network management and routing functions. According to the structure and function, it can be divided into unmanaged, stacked and chassis hubs. From the perspective of LAN application, it can be divided into five categories: single relay, multi-segment, port switching, etc. Its hardware characteristics determine its basic connection role in the early network architecture.
Please open Telegram to view this post
VIEW IN TELEGRAM
网络集线器的主要功能是对接收到的信号进行再生整形放大，以扩大网络的传输距离，同时把所有节点集中在以它为中心的节点上。它工作于OSI（开放系统互联参考模型）参考模型第一层，即“物理层”。集线器与网卡、网线等传输介质一样，属于局域网中的基础设备，采用CSMA/CD（一种检测协议）访问方式。
The main function of a network hub is to regenerate, shape and amplify the received signal to expand the transmission distance of the network, and at the same time concentrate all nodes on the node centered on it. It works at the first layer of the OSI (Open System Interconnection Reference Model) reference model, namely the "physical layer". Like transmission media such as network cards and network cables, hubs are basic devices in local area networks and use CSMA/CD (a detection protocol) access method.
网络集线器属于纯硬件网络底层设备，基本上不具有类似于交换机的“智能记忆”能力和“学习”能力。它也不具备交换机所具有的MAC地址表，所以它发送数据时都是没有针对性的，而是采用广播方式发送。也就是说当它要向某节点发送数据时，不是直接把数据发送到目的节点，而是把数据包发送到与集线器相连的所有节点。
A network hub is a pure hardware network bottom-level device, and basically does not have the "intelligent memory" and "learning" capabilities similar to switches. It also does not have the MAC address table that switches have, so it does not send data in a targeted manner, but in a broadcast manner. That is to say, when it wants to send data to a certain node, it does not send the data directly to the destination node, but sends the data packet to all nodes connected to the hub.