[Bug]: there are some unexpected r-w confilict during tpcc logrunning test and insert-delete oltp test

[aressu1985]

Is there an existing issue for the same bug?

• [X] I have checked the existing issues.

Environment

- Version or commit-id (e.g. v0.1.0 or 8b23a93):65bdd570cb0580e82ef997e72c411ae7a1cd1b6c
- Hardware parameters:
- OS type:
- Others:

Actual Behavior

When run oltp insert-delte and tpcc long running test, there ares some unexpected r-w confilict error .

Expected Behavior

this exception should not be ocurred.

Steps to Reproduce

No response

Additional information

No response

[aressu1985]

Has discussed with @XuPeng-SH , and he know the reason ,and do not need the mo log. pls assign this issue to @XuPeng-SH

[triump2020]

blocked by other S-1 issue.

[triump2020]

no progress

[triump2020]

blocked by other s-1 bug

[triump2020]

blocked by range/deletes optimization

[sukki37]

repro in commit regression test:

https://github.com/matrixorigin/matrixone/actions/runs/5067549359/jobs/9098992700

https://github.com/matrixorigin/matrixone/actions/runs/5067981412/jobs/9100000169

[triump2020]

no progress

[aressu1985]

repro in commit regression test: https://github.com/matrixorigin/matrixone/actions/runs/5130331539/jobs/9229605103

mo log is in s3 s3://mo-checkin-regression/result/9741/

[triump2020]

 reproduced, try to find the cause now.

[triump2020]

in testing

[triump2020]

@aressu1985 pls test.

[aressu1985]

fixed

[aressu1985]

this happen again on commit: 70cd870

log can be get from loki: ns: mo-nightly-regression

[XuPeng-SH]

It should be fixed by #10133

[aressu1985]

this issue still exist https://github.com/matrixorigin/matrixone/actions/runs/5317767539/jobs/9628685537

[sukki37]

repro:https://github.com/matrixorigin/matrixone/actions/runs/5325558080/jobs/9648684647

[triump2020]

blocked

[triump2020]

All the reasons lead to rw conflcit are as follows:

[triump2020]

no progress

[triump2020]

no progress

[triump2020]

on vacation

[triump2020]

blocked

[triump2020]

blocked

[triump2020]

All the reasons lead to rw conflcit are as follows:

@aressu1985 @XuPeng-SH All the causes to RW conflicts are as above, i think it's difficult to fix it complety in V1.0, but fortunately it occurs with a low probability now. @aressu1985 pls downgrade it

[aressu1985]

need to be discussed on next week meeting to confirm how to deal with this issue

[aressu1985]

@XuPeng-SH pls reassign this issue

[XuPeng-SH]

r-w 目前在一些场景下会有概率出现，无法彻底解决，到1.2我们会优化一下流程，后续再彻底避免这个问题

[XuPeng-SH]

not repro any more