After the Spring Festival holiday closed and tackling key problems, we completed freetSDB-V0.1.1 (github.com/freetsdb/fr…) Developed and released on the last day of February!

Following Freetsdb-v0.0.2, the distributed capability is completely unlocked (according to the different DATA characteristics of META node and DATA node, two different distributed models are designed, CP mode of META node, AP mode of DATA node, and support horizontal expansion, custom consistency level, etc.). In V0.1.1, We aligned the performance and ecology InfluxDB.

FreeTSDB’s performance is consistent with that of InfluxDB, and it is, like InfluxDB, far ahead of the other third-party InfluxDB clusters. For example, based on public data, we compared FreeTSDB/InfluxDB with Ali Cloud’s InfluxDB cloud services.

In Ali Cloud ECS (16 vCPU 32 GiB, Ubuntu 18.04 64bit), the write performance of FreeTSDB is consistent with the InfluxDB, 386985.67 point/ SEC, about the same as that of ali Cloud’s InfluxDB service (16 core 64G: ~ 150,000 point/ SEC) 2.58 times.

Many of you may have noticed that “FreeTSDB has almost completed its major features as an open source alternative to The InfluxDB enterprise edition”, and are also curious about the future plans of FreeTSDB, which is what I would like to discuss with you next.

As the saying goes, “If you do not forget why you started out, you will always succeed”. So what is FreeTSDB’s original goal?

The original aspiration of FreeTSDB is to “maximize distributed, high performance and high concurrency”. As a technical person, FreeTSDB was started as a supporting program for InfluxDB Principles and Combat, carrying our technical ideals and technical expression.

We believe that technology should not just be a pile of words, that is to say, we firmly oppose and do not recognize PPT as a product.

We think of words as an expression, and code as an expression, but some things can only be expressed through code, and words alone can’t. We chose the extremely complex scenario of “massive data, real-time, database” as the “farmland”, expressing technology, ideas and thinking through code, distributed, high performance, high concurrency, architecture design, real-time computing, storage engine, low latency query, etc.

So, we would then change the technology stack, and reconstructing the storage engine core modules, such as the target line core indexes such as reading, writing and time series can be orders of magnitude, and high performance, high availability of FreeTSDB real-time computing, storage engines, such as core competence, extend beyond the time-series data of data scenarios, supports a richer system form, Enable more scenarios and businesses.

In addition, I would like to say that our original intention is ideal, but practical and pragmatic is our goal. We hope to solve the challenge of massive timing data scenarios and help you build a real real-time monitoring system and timing data center.

For example, my first contact with InfluxDB was also to solve the problems of actual scenarios, build the central platform for massive timing data, and realize real-time and efficient access computing capability, low-cost storage capability, and low-delay query capability.

When we communicate with people on the Internet, we also receive a lot of requests from actual scenes, such as Ali’s children’s boots:

I do monitoring, selection cost first, no compression time can not buy

Expect an official aligned distributed version

We have also tested the high availability version of XXX, the performance is not good, not as good as standalone, so we have no choice but to separate the library and table, but the middle layer for half a year

This distribution is really hardcore technology, look forward to the beta soon, I will be on it

For example, questions and needs from the industrial Internet scene:

How does the INFLUxDB database communicate with the PLC? Or how to use the influxDB database to monitor PLC working status in real time?

Addressing these issues and supporting requirements will also be a priority in the future roadmap of FreeTSDB.

So, going forward, FreeTSDB will be more open, connecting other open ecosystems and systems while accommodating the InfluxDB ecosystem. We will continue to work deeply in the ecology and scene, serve the needs of DevOps, IoT and industrial Internet scene, and solve the problems of the actual scene realistically and down to earth, enabling business!

Finally, if you are facing pain points and challenges in temporal data scenarios (DevOps, IoT, industrial Internet, etc.), including but not limited to real-time computing, efficient storage, low latency queries, OLAP, etc., please feel free to contact us. I hope FreeTSDB can help you!

Read more:

“Why is InfluxDB | write in the publication of the” InfluxDB principle and actual combat, “”

I want to Talk to you: FreeTSDB is not just an InfluxDB cluster

Welcome to exchange and discuss:

Influxdb-dev

FreeTSDB Technology Exchange Group (QQ) : 663274123