> For example, why would minhittime=0 end all throttling entirely? It
> should only short-circuit the MINHITTIME counter, correct? But with
> 500 I get lots, with 0 I get none.

Yes, 0 would disable it.  Looking at the logic that would indeed cause 
it not to throttle anything.

> Also, I might re-ask from above:
>
> It also doesn't seem to do much anyway. I opened 20 tabs quickly and
> then some were blocked but then the next ones weren't, so what good is
> it? How is it actually throttling someone? Abusers won't care about
> retry-after. Why isn't there a blockperiod attribute? Or how should I
> do that?


You realize you could easily do your own throttle implementation using a 
CFC --- there is nothing really magically going on under the covers.

It was based off the mod_throttle apache module a __long__ time ago, and 
it was pretty much a java port of that logic.   So chances are, it may 
need revisiting.

AGE is milliseconds.


but if you are relying on this feature, i would recommend __NOT__ 
throttling at the application server layer.   by the time you are making 
the decision to throttle here then its probably too late, you've 
consumed resources.

Instead, i would look at putting saying lighttpd, nginx infront of your 
application server and let them do it.

That is why the tag was never updated --- in the 8 odd years since this 
tag was developed, other technologies have moved on to do the job better.

-- 
Open BlueDragon Public Mailing List
 http://www.openbluedragon.org/   http://twitter.com/OpenBlueDragon
 mailing list - http://groups.google.com/group/openbd?hl=en

 !! save a network - please trim replies before posting !!

Reply via email to