maria-developers team mailing list archive
-
maria-developers team
-
Mailing list archive
-
Message #10303
Re: [Commits] 98b2a9c: MDEV-4774 Strangeness with max_binlog_stmt_cache_size Settings
Hi Sachin,
On Fri, Jan 20, 2017 at 3:04 AM, Sachin Setiya <sachin.setiya@xxxxxxxxxxx>
wrote:
> Hi Nirbhay!
>
> On Fri, Jan 20, 2017 at 7:09 AM, Nirbhay Choubey <nirbhay@xxxxxxxxxxx>
> wrote:
>
>> Hi Sachin,
>>
>> The overall patch looks ok. I, however, have a few minor comments inline.
>>
>>
>> On Thu, Jan 19, 2017 at 1:21 AM, SachinSetiya <sachin.setiya@xxxxxxxxxxx>
>> wrote:
>>
>>> revision-id: 98b2a9c967a5eaa1f99bb3ef229ff2af62018ffe
>>> (mariadb-10.0.28-34-g98b2a9c)
>>> parent(s): 9bf92706d19761722b46d66a671734466cb6e98e
>>> author: Sachin Setiya
>>> committer: Sachin Setiya
>>> timestamp: 2017-01-19 11:50:59 +0530
>>> message:
>>>
>>> MDEV-4774 Strangeness with max_binlog_stmt_cache_size Settings
>>>
>>> Problem:- When setting max_binlog_stmt_cache_size=18446744073709547520
>>> from either command line or .cnf file, server fails to start.
>>>
>>> Solution:- Added one more function eval_num_suffix_ull , which uses
>>> strtoull to get unsigned ulonglong from string. And getopt_ull calles
>>> this
>>>
>>
.. cut ..
> diff --git a/mysql-test/suite/binlog/t/binlog_max_binlog_stmt_cache_size.test
>>> b/mysql-test/suite/binlog/t/binlog_max_binlog_stmt_cache_size.test
>>> new file mode 100644
>>> index 0000000..bc30b48
>>> --- /dev/null
>>> +++ b/mysql-test/suite/binlog/t/binlog_max_binlog_stmt_cache_size.test
>>> @@ -0,0 +1,11 @@
>>> +source include/have_log_bin.inc;
>>> +select @@max_binlog_stmt_cache_size;
>>> +
>>> +--let $cache_size=`select @@max_binlog_stmt_cache_size;`
>>> +
>>> +set global max_binlog_stmt_cache_size= 18446744073709547520;
>>> +select @@max_binlog_stmt_cache_size;
>>> +
>>> +set global max_binlog_stmt_cache_size= 18446744073709547519;
>>> +select @@max_binlog_stmt_cache_size;
>>>
>>
>> I would also add tests for ULLONG_MAX and ULLONG_MAX +/- 1.
>>
>> Added the test for ULLONG_MAX+1, I am already testing for ULLONG_MAX and
> ULLONG_MAX-1.
>
Not exactly. What you currently test is the maximum allowed value
'18446744073709547520'
for this variable (which is good). But, I would adittionally like to have
it ULL_MAX (& +/1) which
is 18446744073709551615.
Considering this change, I have no more remarks.
Best,
Nirbhay
Follow ups
References