Details
-
Task
-
Status: Open (View Workflow)
-
Minor
-
Resolution: Unresolved
-
None
-
None
Description
I have got following Table
|
create Table: |
CREATE TABLE `mytable` ( |
`id` int(10) unsigned NOT NULL AUTO_INCREMENT, |
`title` varchar(255) NOT NULL DEFAULT '', |
`icon` blob,
|
PRIMARY KEY (`id`) |
) ENGINE=aria DEFAULT CHARSET=latin1 |
If i do requests on this table the blob field named icon is always read even so i don't request its contents which makes the querys really slow.
i just do a
SELECT id,title FROM mytable |
all the blob fields are read from disk.
If i do it like
SELECT id,title FROM mytable where id in (....) |
all blob fields of all specified rows are read.
I looked at the internet before and found the following claim that blob support in Aria ist better than in MyISAM.
https://mariadb.com/kb/en/mariadb/aria-faq/
-What are the 'design goals' in Aria?
-Better blob handling (than is currently offered in MyISAM, at a minimum).
-No memory copying or extra memory used for blobs on insert/update.
-Blobs allocated in big sequential blocks - Less fragmentation over time.
-Blobs are stored so that Aria can easily be extended to have access to any part of a blob with a single fetch in the future.
Has nothing of the above done yet?
I tried InnoDB but that ended in a total database server crash with a lot of data corruption in the database itself, i was not able to recover my test data. I could crash the server with just two simple SQL statements (tested on mysql).