On Mon, 2005-09-26 at 08:32 +0100, vikas gupta wrote:
> hi Sebastien ,
>
> I am having yet another query related to AIO Support
>
> 1)As you mentioned that aio posix support is provided
> by glibc also so can you tell me on the performance
> basis which one is better (glibc implementation or
> libposix implementation ) and how do we measure the
> performance
Have a look at:
http://www.bullopensource.org/posix/Bench/sysbench-oltp/sysbench.html
for benchmarks using Sysbench and MySQL.
>
> I have compiled glibc with following command
>
> gcc -g $1.c -o $1 -lrt -lpthread
Then you're using glibc AIO implementation based on helper threads
and without taking advantage of kernel support.
>
>
> 2)What posix features is supported by bare kernel and
> libposix implementation, without applying the
> patches.I have broken down my query in following
> parts
Without any of the patches:
>
> 2.1) aio_read/aio_write is supported but what
> limitation are there
Supported but without notification (SIGEV_NONE only).
>
> 2.2) aio_fsync is supported or not
Supported only if the underlying fs implements it
>
> 2.3) what are the limitation with lio_listio
Not supported wihtout the patches.
>
> 2.4) what are the additional feature it provides for
> aio_cancel implementation
>
Needs support from the underlying fs.
>
> 3) Is glibc implementation is providing all the above
> mentioned fetures
Yes, I think so.
>
> 4) Is there any test program that can measure
> efficiency for both glibc and libposix implementation
I personally use Sysbench and have compiled 3 MySQL servers,
one with librt AIO, one with libposix-aio and one with MySQL
native simulated AIO.
You may also try iozone.
>
>
> With Thanks in advance
> Vikas
Hope this helps,
Sébastien.
--
------------------------------------------------------
Sébastien Dugué BULL/FREC:B1-247
phone: (+33) 476 29 77 70 Bullcom: 229-7770
mailto:[email protected]
Linux POSIX AIO: http://www.bullopensource.org/posix
------------------------------------------------------
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
[Index of Archives]
[Kernel Newbies]
[Netfilter]
[Bugtraq]
[Photo]
[Gimp]
[Yosemite News]
[MIPS Linux]
[ARM Linux]
[Linux Security]
[Linux RAID]
[Video 4 Linux]
[Linux for the blind]
|
|