[Disksim-users] Saturation in DISKSIM_malloc()

[Zhe Zhang]

Hi Lanyue,

Thanks for the reply. In my copy of disksim the MAX_QUEUE_LENGTH
is 10,000 instead of 100,000. No wonder I got a lot of saturation errors
=(

It seems my original question was wrong. The saturation is not about
DISKSIM_Malloc().

On 10/25/07, Lanyue Lu <ll2 at rice.edu> wrote:
>
> Hi,
>
> In disksim_logorg.c, there is a macro: MAX_QUEUE_LENGTH 100000, which is the limit of the number of outstanding
> request in the system.
>
> If you request rate is too high, after some time, the number of outstanding requests will be bigger than 100000,
> then you get the error.
>
> The source code related to this is in the function: "static void logorg_mapreqeust_update_stats() " in disksim_logorg.c.
>
> Lanyue Lu
>
> Rice University
>
>
>
> -------------------------------------------------------------
> 发件人：Zhe Zhang
> 发送日期：2007-10-25 16:01:09
> 收件人：disksim-users at ece.cmu.edu
> 抄送：
> 主题：[Disksim-users] Saturation in DISKSIM_malloc()
>
> Hi,
>
> I'm using DiskSim-2.0 with the "Seagate Cheetah 9LP" disk listed on
> the DiskSim website.
> It works fine with some traces, but for some other more intensive
> traces (intervals b/w
> requests are short), it gives me a lot of "Stopping simulation because
> of saturation" error
> messages, and the simulation finally stops running.
>
> I tried allocating more space for disksim by using a larger number for
> the 2nd parameter
> of the function disksim_interface_initialize(disk, len, paramFile,
> outFile), but still the saturation
> errors will kill the simulation. The difference is, if I use larger
> len value, more saturation error
> messages will be printed before the simulation is stopped.
>
> Thanks for any help/suggestions.
>
> Zhe Zhang
> _______________________________________________
> Disksim-users mailing list
> Disksim-users at ece.cmu.edu
> https://sos.ece.cmu.edu/mailman/listinfo/disksim-users
>
>