Ubuntu already improved their boot time

a-aa a-aa at hollowtube.mine.nu
Thu Sep 1 01:43:25 UTC 2005


Erich Schubert wrote:

>Hi,
>  
>
>>Well filtering that out can be done quite easily in a pure C daemon.
>>    
>>
>Only if you can really handle the load. When you get dropped messages
>you can't just filter out what you want, you still might have lost
>some.
>  
>
inotify had the same problem, it will generate a massive amount of hits
(in auditing you can limit to exec for example, in inotify every
read/stat will trigger a hit), with a quick tweak of my database I took
the cpu usage of that system down to near none (time'd it to 0.020,
which is acceptable when I had it running for a while)).

>  
>
>>well, not really, if you scan through memory for \n, replace it with \0,
>>do the system call and add +1 to the buffer you'll scan.  And then
>>simply do a break on \0 found instead of \n.
>>    
>>
>
>On a read-only mmap? I don't have a buffer! (well, I do, thats why I
>have one memcpy - I copy the filename to my buffer and append a \0)
>  
>
hm, right... I had that exact system working with mmap, and tweaked, but
I can't for the life of me remember how it was set up :p

>best regards,
>Erich Schubert
>--
>    erich@(mucl.de|debian.org)      --      GPG Key ID: 4B3A135C    (o_
>  To understand recursion you first need to understand recursion.   //\
>  Wo befreundete Wege zusammenlaufen, da sieht die ganze Welt für   V_/_
>        eine Stunde wie eine Heimat aus. --- Herrmann Hesse
>
>_______________________________________________
>initscripts-ng-devel mailing list
>initscripts-ng-devel at lists.alioth.debian.org
>http://lists.alioth.debian.org/mailman/listinfo/initscripts-ng-devel
>
>  
>




More information about the initscripts-ng-devel mailing list