On Tue, Aug 19, 2014 at 10:38 AM, Richard Hipp <d...@sqlite.org> wrote:

>
>
>
> On Tue, Aug 19, 2014 at 10:34 AM, Levente Kovacs <leventel...@gmail.com>
> wrote:
>
>> I started to worry about this issue, because I am in a middle of an
>> application development, and yesterday, it started to work, and I only
>> SELECT
>> a few times, and it makes a noticeable disk access. I'm still on magnetic
>> HDD,
>> but the application will be running on SSD or Flash drive.
>>
>> Let me start again by a stupid question. If I do a SELECT... does this
>> performs any disk write operation?
>>
>
> No.  SQLite doesn't.  But your filesystem might decide to update the mtime
> on an inode.
>

Correction:  If you do a sufficiently complex SELECT, SQLite might need to
manifest a view or do a large sort that won't fit in RAM.  In those cases
it will write to disk.  You can disable that using PRAGMA
temp_store=MEMORY.  http://www.sqlite.org/pragma.html#pragma_temp_store

Additional information:  http://www.sqlite.org/tempfiles.html



>
>
>>
>> I know that sqlite can store the database in RAM. Is there any way to
>> (periodically) write the database to a regular sqlite file?
>>
>
> http://www.sqlite.org/backup.html
>
>
>
>>
>> Does sqlite calls 'sync()' after an UPDATE or INSERT?
>>
>
> http://www.sqlite.org/atomiccommit.html
>
> --
> D. Richard Hipp
> d...@sqlite.org
>



-- 
D. Richard Hipp
d...@sqlite.org
_______________________________________________
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users

Reply via email to