(Author of Badger here) First of all, thanks to the growing Go community for using Badger. Within a short period of time since the original announcement [-1],

(Author of Badger here) First of all, thanks to the growing Go community for using Badger. Within a short period of time since the original announcement [-1], Badger has become very popular. A lot has changed since the original post. Despite being tagged as a non-goal at the time of release, Badger now runs concurrent ACID transactions providing serializable snapshot isolation guarantees. In fact, Badger is being used to serve Petabytes of data (yes, PB, not TB). On my long TODO list to write a blog post about it. It has stood the test of time, being directly used by the community and Dgraph [0]. We have put it through many different crash testing scenarios, including ALICE[1], and run Jepsen style bank transaction tests every night for 8h [2]. There're some common questions around benchmarks on our blog posts. They were done before Badger had transactions and are somewhat old by now. But, Badger's read-write performance is still spectacular. In terms of write throughput, it outperforms RocksDB (called from Go), LevelDB BoltDB, and in terms of read performance, it goes neck-to-neck against B+ tree based DBs, which typically perform better than LSM trees. Recently, there was a benchmark by Ethereum 2.0 folks, which I ran [3]. You can see the ASCIInema here [4]. There's an open bounty of $1337 on Badger to help us find scenarios where Badger could lose data [5]. So, if you got ideas, throw them in! Thanks again for using Badger! [-1]: Well, the original link here on HN [0]: https://dgraph.io https://dgraph.io [1]: https://blog.dgraph.io/post/alice/ https://blog.dgraph.io/post/alice/ [2]: https://github.com/dgraph-io/badger/commit/c10276c9d3b0c42744ed41b9177f4a9f021d8e08#diff-a3128b018385dad1c7c20cb322b8267b https://github.com/dgraph-io/badger/commit/c10276c9d3b0c4274... [3]: https://github.com/rawfalafel/db-benchmarks/pull/1 https://github.com/rawfalafel/db-benchmarks/pull/1 [4]: https://asciinema.org/a/Qgyf8C50YRbtH7WpixSaBKDUQ https://asciinema.org/a/Qgyf8C50YRbtH7WpixSaBKDUQ [5]: https://github.com/dgraph-io/badger/issues/601 https://github.com/dgraph-io/badger/issues/601