Bug #1827 Replication is severly delayed

• Submitted: 13 Nov 2003 4:28
• Modified: 1 Jan 2004 14:31
• Reporter: Marek Les
• Status: No Feedback
• Category: MySQL Server: Replication
• Severity: S2 (Serious)
• Version: 4.0.16
• OS: Linux (Linux)
• Assigned to: Assigned Account
• CPU Architecture: Any

[13 Nov 2003 4:28] Marek Les

Description:
On some occasions, my replicated slave encouteres a delay which I believe may be causes by a bug in MySQL. The replication stops for several minutes (10 min at maximum) and then starts again. I found out that it happens when there's a "Waiting for release of readlock" line in the MySQL process list. It looks like this:

| 1      | system user |           |      | Connect | 779340 | Waiting for master to send event |                                                                                                      |
| 2      | system user |           | ubox | Connect | 857    | Waiting for release of readlock  | UPDATE user_email SET last_used = NOW(), last_client_ip = 'xxx.175.36.79' WHERE reg_user_id = 279784 |
| 580178 | root        | localhost | ubox | Query   | 212    | Writing to net                   | SELECT /*!40001 SQL_NO_CACHE */ * FROM `reg_user_data`                                               |

But there's no process accesing the user_email table. What could be the problem?

How to repeat:
Don't know.

[1 Dec 2003 14:31] Dean Ellis

Check that you do not have a backup of some sort running (via mysqldump for instance) that is acquiring read locks on the table.  Otherwise, use mysqlbinlog and see what other statements may be causing the lock.

I am, otherwise, unable to repeat this behavior.

Thank you.

[14 Feb 2005 22:54] Bugs System

No feedback was provided for this bug for over a month, so it is
being suspended automatically. If you are able to provide the
information that was originally requested, please do so and change
the status of the bug back to "Open".