roadmap: add TiKV 2024 roadmap #16924
Conversation
Signed-off-by: zhangjinpeng87 <zzzhangjinpeng@gmail.com>
|
[REVIEW NOTIFICATION] This pull request has not been approved. To complete the pull request process, please ask the reviewers in the list to review by filling The full list of commands accepted by this bot can be found here. Reviewer can indicate their review by submitting an approval review. |
zhangjinpeng87
requested review from
glorv,
tonyxuqqi,
Connor1996 and
LykxSassinator
Signed-off-by: zhangjinpeng87 <zzzhangjinpeng@gmail.com>
Signed-off-by: zhangjinpeng87 <zzzhangjinpeng@gmail.com>
ti-chi-bot
bot
added
release-note-none
and removed
do-not-merge/release-note-label-needed
labels
Signed-off-by: zhangjinpeng87 <zzzhangjinpeng@gmail.com>
|
@easonn7 This PR will be public for at least 7 days for reviewing and updating. |
Co-authored-by: Bisheng Huang <hbisheng@gmail.com> Signed-off-by: Jinpeng Zhang <zzzhangjinpeng@gmail.com>
Co-authored-by: Bisheng Huang <hbisheng@gmail.com> Signed-off-by: Jinpeng Zhang <zzzhangjinpeng@gmail.com>
Co-authored-by: Bisheng Huang <hbisheng@gmail.com> Signed-off-by: Jinpeng Zhang <zzzhangjinpeng@gmail.com>
Co-authored-by: Bisheng Huang <hbisheng@gmail.com> Signed-off-by: Jinpeng Zhang <zzzhangjinpeng@gmail.com>
Co-authored-by: Bisheng Huang <hbisheng@gmail.com> Signed-off-by: Jinpeng Zhang <zzzhangjinpeng@gmail.com>
Signed-off-by: zhangjinpeng87 <zzzhangjinpeng@gmail.com>
|
|
||
| ### RocksDB Engine | ||
|
|
||
| Newer version RocksDB introduced some promising features like "WAL compression support" and "async-io", we are going to upgrade TiKV's RocksDB to 8.10 version and leverage these powerful features. |
There was a problem hiding this comment.
Maybe also explore using async rocksdb API which integrates better with TiKV's read pool.
Co-authored-by: glorv <glorvs@163.com> Signed-off-by: Jinpeng Zhang <zzzhangjinpeng@gmail.com>
|
|
||
| TiKV has some soft limits for different components' memory usage like scheduler, raftstore, but there are still some components like coprocessor don't have memory quota limit. In the coprocessor, we use the amount of pending queries to throttle the resources usage of coprocessor, but this still can result in Out-Of-Memory(OOM) when there are too many "BIG" queries which retrieve a large amount of data. | ||
|
|
||
| ## Powerful and Robust Storage Engine |
There was a problem hiding this comment.
May be add experimenting with using io_uring based https://github.com/bytedance/monoio to replace sync IO for raft engine.
|
|
||
| This domain aims to enhance TiKV to provide predictable single-digit millisecond latency under different scenarios like too many MVCC versions, disk IO temporarily jitter, high read QPS in a small data range, etc. | ||
|
|
||
| ### In-Memory Engine |
There was a problem hiding this comment.
would it make sense to add a comment about other uses for in-memory usage?
|
Is current performance good enough? |
What is changed and how it works?
Issue Number: Close #16923
What's Changed: added TiKV 2024 updated roadmap, the major investing directions are:
Notice: this PR will be public for at least 7 days for reviewing and updating.
Related changes
pingcap/docs/pingcap/docs-cn:Check List
Tests
Side effects
Release note