neuyilan commented on issue #3954: URL: https://github.com/apache/iotdb/issues/3954#issuecomment-920542383
I think the current focus is on the following two points: 1. Can we implement a good and correct raft framework. The answer must be **YES**, but **WHEN**? One year or two? With the development of other time-series databases, such as `influxdb_ iox`、`tdengine`、`matrixdb`. Their market share is increasing. As an industrial product, if they want to have their own competitiveness, they must have the characteristics of rapid iteration and stability. However, over the past year, there is no industrial version of cluster version of `IoTDB`. From this perspective, I prefer to build a usable and secure distributed `IoTDB` version first. No matter who uses the raft framework, and I think it is a faster way to integrate one that has been widely used in the industry than to implement it yourself. 2. How good is the performance of the raft framework implemented by yourself? As far as we know, the performance improvement of our cluster version is as follows: 1. async apply； 2. The query does not follow the raft, and the consistency of the query can be determined according to the configuration. But as far as I know, the apply function of `Apache Ratis `framework can be decided by users. Moreover, the read request of Ratis does not follow the raft, so I don't think the performance will degrade much. 3. What is a competitive product? A safe, stable and widely used product is a competitive product. If there are no available products for users, what about competitiveness? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
