List:Commits« Previous MessageNext Message »
From:Sven Sandberg Date:April 9 2008 8:34am
Subject:Re: bk commit into 5.1 tree (skozlov:1.2552)
View as plain text  
Serge Kozlov wrote:
> Sven Sandberg wrote:
>> Hi Serge,
>>
>> Thank you for your simplifications. This looks much better! I had a
>> couple of comments on my last review which you did not react to - see 
>> below.
> 
> OK. You mean adding to header info from WL

Almost. The worklog contains alternative suggestions which you do not 
implement and hence should not be described in the comment. I think that 
you only use LLD 1.2, so LLD 1.1, 1.3 should not be described.

I also think the comment should be more explicit than LLD 1.2: instead 
of just saying "etc", it should list all tables and triggers on master, 
and then it should say why the slave's sql thread stops with an error 
when the table is gone.

Sorry if I sound picky, but good comments are really, really useful when 
the code is used later on.

> (and number) and sync slave 
> with master at end of test. Will do.
> 
> Does it approved with these changes?

Sure, commit a patch and I'll approve after reading it again.

> 
>>
>> Serge Kozlov wrote:
>> [...]
>>> diff -Nrup a/mysql-test/suite/rpl/t/rpl_slave_grp_exec.test 
>>> b/mysql-test/suite/rpl/t/rpl_slave_grp_exec.test
>>> --- /dev/null    Wed Dec 31 16:00:00 196900
>>> +++ b/mysql-test/suite/rpl/t/rpl_slave_grp_exec.test    2008-04-02 
>>> 18:51:11 +04:00
>>> @@ -0,0 +1,164 @@
>>> +#############################################################
>>> +# Author: Serge Kozlov <skozlov@stripped>
>>> +# Date:   03/21/2008
>>> +# Purpose: Testing slave group execution: stop in middle
>>> +# of a group (of events) should be immpossible on slave.
>>
>> Please include an explanation that more precisely explains what should 
>> happen on the slave when it is stopped in the middle of a group.
>>
>>> +# Group of events means set of statement between BEGIN/COMMIT
>>> +# for transactional engines or set of events created by +# invoked 
>>> features (like triggers, sp) for non-transactional +# engines.  
>>
>> Please include a short paragraph with a high-level description of 
>> *how* you accomplish stopping the slave in the middle of a group.
>>
>> Please include the WL number here.
>>
>> [...]
>>> +# Clean up
>>> +--echo *** Clean up ***
>>> +--connection master
>>> +DROP TABLE t1,t2,t3;
>>
>> Please do sync_slave_with_master at end of test.

-- 
Sven Sandberg, Software Engineer
MySQL AB, www.mysql.com
Thread
bk commit into 5.1 tree (skozlov:1.2552)Serge Kozlov2 Apr
  • Re: bk commit into 5.1 tree (skozlov:1.2552)Sven Sandberg3 Apr
    • Re: bk commit into 5.1 tree (skozlov:1.2552)Serge Kozlov3 Apr
      • Re: bk commit into 5.1 tree (skozlov:1.2552)Sven Sandberg9 Apr
  • Re: bk commit into 5.1 tree (skozlov:1.2552)He Zhenxing18 Apr
    • Re: bk commit into 5.1 tree (skozlov:1.2552)Serge Kozlov18 Apr
      • Re: bk commit into 5.1 tree (skozlov:1.2552)He Zhenxing18 Apr
        • Re: bk commit into 5.1 tree (skozlov:1.2552)Serge Kozlov18 Apr
          • Re: bk commit into 5.1 tree (skozlov:1.2552)He Zhenxing19 Apr