Tungsten issues unnecessary queries for autocommit

[GoogleCodeExporter]

In reviewing queries to see why tungsten can't keep up in single threaded mode 
I found this :

  event_time: 2013-05-22 21:54:44
   user_host: tungsten[tungsten] @ 
   thread_id: 12
   server_id: 253134
command_type: Query
    argument: COMMIT
*************************** 278. row ***************************
  event_time: 2013-05-22 21:54:44
   user_host: tungsten[tungsten] @ 
   thread_id: 12
   server_id: 253134
command_type: Query
    argument: select @@autocommit
*************************** 279. row ***************************
  event_time: 2013-05-22 21:54:44
   user_host: tungsten[tungsten] @ 
   thread_id: 12
   server_id: 253134
command_type: Query
    argument: set autocommit=1
*************************** 280. row ***************************
  event_time: 2013-05-22 21:54:44
   user_host: tungsten[tungsten] @ 
   thread_id: 12
   server_id: 253134
command_type: Query
    argument: select @@autocommit
*************************** 281. row ***************************
  event_time: 2013-05-22 21:54:44
   user_host: tungsten[tungsten] @ 
   thread_id: 12
   server_id: 253134
command_type: Query
    argument: set autocommit=0

Original issue reported on code.google.com by winm...@gmail.com on 22 May 2013 at 10:00

[GoogleCodeExporter]

Hi, could you please add information about exactly what you were doing 
including the environment in which you were executing and what type of 
misbehavior you observed?  It is not possible to determine a problem from this 
description.  Thanks!

Original comment by robert.h...@continuent.com on 23 May 2013 at 4:41

[GoogleCodeExporter]

Single threaded replication between two servers. Can you specify what 
additional details we can provide?

Original comment by winm...@gmail.com on 30 May 2013 at 6:22

[GoogleCodeExporter]

Please say which queries you ran on the master that have caused the behavior 
you are observing.

Original comment by g.maxia on 30 May 2013 at 6:30