Bug #82769 Timeout on metadata lock breaks session

• Submitted: 28 Aug 2016 18:55
• Modified: 11 May 2018 19:17
• Reporter: Daniël van Eeden (OCA)
• Status: Can't repeat
• Category: MySQL Server: Document Store: MySQL Shell
• Severity: S3 (Non-critical)
• Version: 1.0.4
• OS: Any
• CPU Architecture: Any

[28 Aug 2016 18:55] Daniël van Eeden

Description:
A metadata timeout breaks the mysql shell session.

How to repeat:
SESSION 1:
mysql [localhost] {msandbox} (test) > flush tables with read lock;
Query OK, 0 rows affected (0.00 sec)

SESSION 2:
mysql-py> db.t1.find()
MySQL Error (2000): Operation canceled
mysql-py> 

SESSION 1:
mysql [localhost] {msandbox} (test) > unlock tables;
Query OK, 0 rows affected (0.00 sec)

SESSION 2:
mysql-py> db.t1.find()
MySQL Error (2000): Bad file descriptor
mysql-py> db.t1.find()
MySQL Error (2000): Bad file descriptor
mysql-py> db.t1.find()
MySQL Error (2000): Bad file descriptor
mysql-py> db.t1.find()
MySQL Error (2000): Bad file descriptor

[29 Aug 2016 8:41] MySQL Verification Team

Hello Daniël,

Thank you for the report.

Thanks,
Umesh

[11 May 2018 19:17] Hema Sridharan

From support verification team:
*No Longer reproducible*
 
Fedora 24x64
 
MySQL Server 5.7
 
21082704.mysql-shell-1.0.6-0.1.fc24.x86_64
 
No problems founds after repeat the steps in both sessions.