No doubt throttling is used for API calls, but IP address throttling  
of the full user interface ought to be managed quite differently. If  
anyone has seen that occur, I would be interested to hear of it.


On Aug 5, 2009, at 2:34 PM, Jon Gorman wrote:

> On Wed, Aug 5, 2009 at 1:05 PM, Eric Hellman<[log in to unmask]> wrote:
>> I doubt that very much. It's very common for corporate sites to  
>> channel all
>> their traffic through gateways. I would assume that google was  
>> smart enough
>> to recognize that your usage pattern was not that of many users  
>> coming from
>> a single IP address, but rather that of a harvesting robot. The two
>> activities have very different log signatures.
>>
>
> Uh, actually, Google has in the past throttled some services based on
> the ip address.  I'm pretty sure it was mentioned before on this list
> and I can verify it myself.  Look for some of Jonathan Rochkind's
> questions about a year ago.  The original api used with GBS seemed
> very prone to this.  I know others hit issues and when our consortium
> tried to use a proxy of the original api due to some technical issues
> they ran into this.  (First couple of hundred hits would be golden,
> the rest just would return http errors).  There's a newer one out
> there now that apparently doesn't use this throttling, but I'm not
> positive of the details.  An organization may still have to warn
> google about it.
>
> There's a reason why the original api strongly encouraged folks to do
> things via a ajaxy call on the client.  I'm guessing part of the
> reason for the "new api" was to address these issues.
>
> Jon Gorman

Eric Hellman
President, Gluejar, Inc.
41 Watchung Plaza, #132
Montclair, NJ 07042
USA

[log in to unmask]
http://go-to-hellman.blogspot.com/