-
Generate a CSV showing the delayed write rate over time.
This should greatly help diagnose issues related to writing to the DB.
-
The new write buffer manager should be with allow_stalls enabled.
This is also relevant for the Speedb tuning function
-
### Search before asking
- [X] I had searched in the [issues](https://github.com/apache/kvrocks/issues) and found no similar issues.
### Motivation
We're now supporting the speedb and rocksdb as i…
-
Currently, the speedb rust binding relies on the rust rocksdb rust binding (https://github.com/rust-rocksdb/rust-rocksdb) and therefore currently does not expose many configurations. The most importan…
-
- Report the following per block cache:
- Per CF sharing the block cache in the DB, report the following information for the data blocks, index blocks, filter blocks, and compression dictionary blo…
-
## Current Default
- `target_file_size_multiplier = 1`
- `block_size = 4096`
- `OptimizeLevelStyleCompaction(512M)` implies
- `target_file_size_base = 64M`
- snappy/lz4 compression types
…
-
**Describe the bug**
setting a WBM for a single db with db_write_buffer_size, causes a WBM to be initialized in SanitizeOptions with the default value of initiate_flushes which is true.
the Proa…
-
As part of #488 there was a need to sanitize a block based table option (specifically, the optimize_filters_for_memory).
In addition, the flag should be sanitized only in a specific plug-in (Speedb's…
-
1 每个命令生成一个 shared_ptr:pikiwidb 不复制命令对象,使用 context 优化,发包池化;
2 buffer 池化,省掉一些 malloc;
3 命令发送 -> 处理 -> 多线程:缓存下命令;StringView
4 info 命令,admin 拆出来,不要放在主线程;
5 io worker 调度不均衡,把任务抽象;
6 可观测:命令耗时,worker 调度…
-
i dont see a reason why a user should set max simultaneous flushes for the proactive WBM while theres already a setting for the max bg flushes a group of dbs can use.
seems logical to me that this…