matrixorigin
[Bug]: CN lost connection
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):
- Hardware parameters:
- OS type:
- Others:
Actual Behavior
Running same test in #9646, but with a better machine (32G to 64G memory), you will see cn lost connection.
ERROR 20503 (HY000) at line 40: stream closed
Expected Behavior
No response
Steps to Reproduce
No response
Additional information
No response
in local test:
create database if not exists db1;
use db1
drop table if exists t;
create table t (i int, j int);
insert into t values (1, 1), (2, 2), (3, 3), (4, 4), (5, null), (null, 5);
insert into t select * from t;
insert into t select * from t;
insert into t select * from t;
insert into t select * from t;
insert into t select * from t;
select count(*) from t;
insert into t select * from t;
insert into t select * from t;
insert into t select * from t;
insert into t select * from t;
insert into t select * from t;
select count(*) from t;
insert into t select * from t;
insert into t select * from t;
insert into t select * from t;
insert into t select * from t;
insert into t select * from t;
select count(*) from t;
insert into t select * from t;
insert into t select * from t;
insert into t select * from t;
insert into t select * from t;
insert into t select * from t;
select count(*) from t;
insert into t select * from t;
insert into t select * from t;
insert into t select * from t;
insert into t select * from t;
insert into t select * from t;
select count(*) from t;
delete from t where i = 1;
delete from t where i = 2;
select count(*) from t;
insert into t select * from t;
select count(*) from t;
above script will result in: 2023/05/24 12:21:17.672100 +0800 ERROR logservicedriver/appender.go:70 append failed: internal error: message body 116720140 is too large, max is 104857600
该问题的分析见https://github.com/matrixorigin/matrixone/issues/9447#issuecomment-1576675658
后续由 @triump2020 再做一些优化。
ERROR 20503 (HY000) at line 40: stream closed
出现该错误或者其他的连接关闭的错误的原因是，rpc框架中的gc任务会检查每个连接的活跃状态，当超过一定时间（默认1分钟）没有数据时，就会关闭这个连接。
delete语句的commit时间长，导致其连接被gc给close了，所以执行失败，需要优化。
#10418 cannot fix this issue. This depends on relatively significant refactoring.
depends on https://github.com/matrixorigin/matrixone/issues/11805 https://github.com/matrixorigin/matrixone/issues/11804