Re: Amazing with MySQL 5.1.7

[John David Duncan]

Also -- maybe it is more informative to include Index_length in the  
calculation...



On Mar 12, 2006, at 7:38 PM, John David Duncan wrote:

> Hi!
>
> In 5.0, NDB has fixed-length rows, so VARCHAR is really CHAR.
> In 5.1, NDB has real VARCHARs.
>
> Maybe that explains the difference?
>
> JD
>
>
> On Mar 12, 2006, at 7:16 PM, David wrote:
>
>> I've posted a question in http://lists.mysql.com/cluster/3237
>> I have a table with MyISAM store engine and its size is about 340M  
>> with
>> 5,000,000 rows
>> When I convert it to NDBCLUSTER with 5.0.18 store engine, its size  
>> becomes
>> 1.4G
>> With MySQL 5.1.7 beta, its size becomes 259.84M
>> Huh, much smaller than MySQL 5.0.18
>> But I confused about it, why its smaller then the original ?
>> As mentioned in MySQL Reference Manual, it has overheads in MySQL  
>> Cluster,so
>> the size is theoretically bigger than original
>> My calculation is:
>> select table_name,engine,round(data_length/1024/1024,2) from
>> Information_Schema.tables where table_name = 'user1'
>> +------------+------------+--------------------------------+
>> | table_name | engine | round(data_length/1024/1024,2) |
>> +------------+------------+--------------------------------+
>> | user1 | MyISAM | 332.81 |
>> | user1 | NDBCLUSTER | 259.84 |
>> +------------+------------+--------------------------------+
>>
>>
>>
>>
>> Thanks, :)
>
>
> -- 
> MySQL Cluster Mailing List
> For list archives: http://lists.mysql.com/cluster
> To unsubscribe:    http://lists.mysql.com/cluster?unsub=1
>