Details
-
Bug
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Fixed
-
13.2
-
None
-
Can result in unexpected behaviour
-
Q4/2026 Server Maintenance, Q1/2027 Server Maintenance
Description
Issue:
------
After connection to remote server is broken due to timeout ,
Subsequent query fails to open new connection ,uses broken connection and errors out
MariaDB [m8]> truncate s1;
ERROR 1296 (HY000): Got error 10000 'Error on remote system: 2006: Server has gone away' from FEDERATED
Testcase:
--------------
MariaDB [m8]> UPDATE s1 a JOIN s1 b SET a.a = 1;truncate s1;truncate s1;
ERROR 1159 (08S01): Got timeout reading communication packets
ERROR 1296 (HY000): Got error 10000 'Error on remote system: 2006: Server has gone away' from FEDERATED
ERROR 1159 (08S01): Got timeout reading communication packets
Issue details:
---------------
Update Query takes 10sec to execute
Say federated server started with net_read_timeout=3
So update query errors in 3sec bcz of timeout
Connection to remote server is broken.
1st Truncate -->Instead of opening new connection to remote server, Tries to re-use broken connection ,hence error's out with "Server has gone away" -->Bug ??
2nd Truncate --Opens new connection as expected.
Version used:
----------------
Debug build
Branch : 13.2-MDEV-41243-fedx-pushdown-assert-for-upd-del-prepared-stmts
commit: e7d83013b31c4cc97a0003a0e58b707d94cba042
Version : 13.1.0
How to repeat:
----------------
Slave:
create table s1( a int); |
Master: Start server with federated engine enabled
Start server with --net_read_timeout=3 |
create table s1(a int) ENGINE="FEDERATED" CONNECTION='mysql://u3:pass1@127.0.0.1:3306/m8/s1'; |
insert into s1 select seq from seq_1_to_1000; |
 |
set global federated_pushdown=ON; |
UPDATE s1 a JOIN s1 b SET a.a = 1;truncate s1; //Notice "server has gone away" error |
Attachments
Issue Links
- is caused by
-
MDEV-39226 Direct multi-table update/delete interface and FederatedX implementation
-
- Closed
-