Shulou(Shulou.com)06/01 Report--

What is the RPC transmission in big data's architecture? in view of this question, this article introduces the corresponding analysis and answer in detail, hoping to help more partners who want to solve this problem to find a more simple and feasible way.

RPC (Remote ProcedureCall), that is, remote procedure call, implements the previous call between two servers through the network, and RPC spans the transport layer and the application layer.

The procedure for calling RPC:

The client initiates a RPC request, and the local call client stub is responsible for serializing the called interfaces, methods and parameters according to the agreed protocol, and then the RPCRuntime instance of the RPC framework is transmitted to the remote server through socket. After receiving the instance, the remote server-side RPCRuntime deserializes the instance through server stub and initiates the final method call.

Excellent RPC open source framework:

Domestic Dubbo, Motan, foreign Thrift, gRPC

Understand the implementation principle of RPC from the following aspects:

1, dynamic agent: agent is to do something, do not have to do it in person, find an agent to do it, dynamic refers to the method and the number of tasks that can be dynamically specified.

2. Reflection: refers to the ability of a computer program to access, detect and modify its own state or behavior when it is running. Pass in the remote service name and method name, automatically locate the method that needs to be called through reflection, and then pass in the input parameters to make the RPC call.

3. Serialization: converts the data body in memory into a byte stream, and deserialization converts a byte stream into a data body.

4. Network programming: RPC is generally based on TCP protocol.

Location method of service RPC request exception:

Step 1, analyze the network situation to see if the delay is serious and whether there is a TCP retransmission.

Step 2, analyze the operation of the server side and the calling side to see if the pressure is high, such as CPU utilization, CPU load, memory footprint, etc.

Step 3, check to see if the transfer object is large and complex, which has a big impact on serialization.

Step 4, if the server has a queue, you can try to reduce the queue, or change to a fixed thread pool to reduce the number of thread pools.

Step 5, control CPU not to be too high, try not to exceed 80%, there is a formula (number of cores * 7) and then go to the percentage, eg:4c container, then 2.8% utilization is normal, when it reaches 80%, there may be thread blocking.

So much for the answer to the question about what RPC transmission is in big data's architecture. I hope the above content can be of some help to you. If you still have a lot of doubts to be solved, you can follow the industry information channel to learn more about it.

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.