Cook device vs. Raw device 
Author Message
 Cook device vs. Raw device

I haven't been following this thread to closely, so pardon me if I am
repeating someone else's observation; but I always thought the most
important difference between raw I/O vs. cooked I/O was data integrity
following a crash.  Cooked I/O might report a change as committed, while in
reality the data has only been committed to the disk cache/buffer.  With raw
I/O, the data is committed to disk by the engine not the OS.

Walt

Quote:
> -----Original Message-----

> Sent:      Wednesday, September 15, 1999 3:02 PM
> To:        William Rice
> Cc:        informix-list
> Subject:   RE: Cook device vs Raw device

> This is a thread that has always interested me.  I have been reading this
> group for awhile, and was concerned when I came to my current place of
> employment
> and discovered they were using filesystems to hold data.  When we
> got the oppurtunity to do some limited testing on raw vs cooked we
> could only find at most a  1.5% difference between informix's performance
> to raw disk vs. cooked.
> We watched the following things:
>   iostat,
>   vmstat
>   wall time

> We were running AIX 4.3.2 on a 520
> Informix 7.31UC2

> Only 2 disk drives, one of which was dedicated to informix data.
> Under 2 Gigabytes of data in the whole database

> These processes were simple queries(no more than three tables joined)
> and the loading of data into tables.

> The longest running process was somewhere around 15 minutes.
> The processes tested caused the system to be heavily io bound for some
> and CPU for others.

> From reading the newsgroup I got the impression that I could count on at
> least
> a 10% difference between cooked and raw, yet the tests I ran did not turn
> out
> that way.

> The data involved was literally identical.  I was curious what factors
> might
> be causing the performance not to increase.

> Will

> ------------------------------------------------------------
> This e-mail has been sent to  you  courtesy of OperaMail,  a
> free  web-based  service  from  Opera  Software,  makers  of
> the award-winning Web Browser - http://www.***.com/
> ------------------------------------------------------------



Wed, 18 Jun 1902 08:00:00 GMT
 Cook device vs. Raw device

I was under the impression that due to the fact the file was opened
 with O_SYNC that data integrity was not a concern.

definition of O_SYNC - Have each write wait for physicl I/O to complete.
(This is not part of POSIX.1 but is supported by SVR4)

Will

Quote:

=====
>I haven't been following this thread to closely, so pardon me if I am
>repeating someone else's observation; but I always thought the most
>important difference between raw I/O vs. cooked I/O was data integrity
>following a crash.  Cooked I/O might report a change as committed, while in
>reality the data has only been committed to the disk cache/buffer.  With raw
>I/O, the data is committed to disk by the engine not the OS.

>Walt

------------------------------------------------------------
This e-mail has been sent to  you  courtesy of OperaMail,  a
free  web-based  service  from  Opera  Software,  makers  of
the award-winning Web Browser - http://www.operasoftware.com
------------------------------------------------------------


Wed, 18 Jun 1902 08:00:00 GMT
 
 [ 2 post ] 

 Relevant Pages 

1. Cook device vs Raw device

2. Cooked files vs raw devices

3. Raw vs. cooked devices

4. Raw Device Vs.Cooked File

5. Raw devices Vs. File system devices

6. raw device or not raw device

7. Informix/AIX block raw device to character raw device

8. Cooked verses raw devices and dsync

9. Raw vrs Cooked device

10. Red Hat 7.2 : Ext3 cooked file system versus raw devices

11. Raw I/O vs files (was Re: Raw partitions / cooked files)

12. DISK DEVICE vs TAPE DEVICE BACKUP


 
Powered by phpBB® Forum Software