ZanRedisDB
ZanRedisDB copied to clipboard
Yet another distributed fault-tolerant key-value database Compatible with Redis written in Golang.
如题
# github.com/youzan/gorocksdb /root/soft/rocksdb/librocksdb.a(db_impl.o): In function `rocksdb::DBImpl::DBImpl(rocksdb::DBOptions const&, std::string const&, bool, bool)': /root/soft/rocksdb/./util/compression.h:275: undefined reference to `ZSTD_versionNumber'
Support using the k8s operator to deploy a full cluster with the PersistentVolume
some patch can be merged from etcd: raft: Introduce CommittedEntries pagination https://github.com/etcd-io/etcd/pull/10063 https://github.com/etcd-io/etcd/pull/9982 https://github.com/etcd-io/etcd/pull/10258 raft: don't allocate slice and sort on every commit https://github.com/etcd-io/etcd/pull/8689 raft: provide protection against unbounded Raft...
hi,感谢开源!我在文档找了下关于数据结构映射的文档,没有太详细的说明,曾经关注在rocksdb模拟redis跳表的实现,特别是*rev* 相关的命令,如: zrevrange、zrevrank、zrevrangebyscore。这些接口在既有的开源项目中实现的非常不理想,毕竟,leveldb的反向迭代器也是比正向慢多了,虽然官网只是提到一句:  我在另个实现中测试了其普通zrange的实现,大概1亿条数据,后来看到其实现是靠正向遍历:  这是完全不能用于生产环境的实现! 所以,zanRedisDB能否说明下一些操作的时间复杂度以及实现呢?