You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Please answer these questions before submitting your issue. Thanks!
What did you do?
There are 20 transactions running insert statement from upstream, each transaction will execute 40 insert statements and each statement contains 10 rows. There are add/drop column options at the same time. The amend transaction is on, so the transactions above can both executed successfully.
When all the DDL options are done, commit the 20 transactions.
After the all the transaction are committed, insert a record into check point table(INSERT INTO check_point_t_114514 VALUES(1605785449)).
Waiting for binlog sync all the data into downstream, we take the existence of 1605785449 in the downstream as the evidence the above transactions are all synced.
Diff the upstream and down stream.
What did you expect to see?
The data of upstream and downstream are same in step 5.
What did you see instead?
Some records are not synced into downstream yet in step5. However SELECT * FROM t_114514 got 8,000 records from upstream, but only 7989 from downstream.
From the log of downstream, there still some replace-into statements, trace the next replace-into statement after writing check point in downstream. This record is written by first line captured from upstream's log, the transaction's startTs is 420947016752300087, also it's successcully committed before we writing check point in upstream.
Please provide the relate downstream type and version of drainer.
(run drainer -V in terminal to get drainer's version)
Release Version: v4.0.8-1-g336b519
Git Commit Hash: 336b51969ce563fc46e9c8d422ef71a1c40d5376
Build TS: 2020-11-19 02:38:44
Go Version: go1.15.2
Go OS/Arch: linux/amd64
The text was updated successfully, but these errors were encountered:
Bug Report
Please answer these questions before submitting your issue. Thanks!
INSERT INTO check_point_t_114514 VALUES(1605785449)
).1605785449
in the downstream as the evidence the above transactions are all synced.The data of upstream and downstream are same in step 5.
Some records are not synced into downstream yet in step5. However
SELECT * FROM t_114514
got 8,000 records from upstream, but only 7989 from downstream.Here are some logs and simple analysis.
Upstream:
Downstream:
From the log of downstream, there still some replace-into statements, trace the next replace-into statement after writing check point in downstream. This record is written by first line captured from upstream's log, the transaction's startTs is 420947016752300087, also it's successcully committed before we writing check point in upstream.
(run
drainer -V
in terminal to get drainer's version)The text was updated successfully, but these errors were encountered: