From: "Daryl C. W. O'Shea"

> jdow wrote:
>> From: "Daryl C. W. O'Shea"
>>> Loren Wilton wrote:
>>>>>>>> Mike Grau 07/09/07 5:15 PM >>>
>>>>> On 07/09/2007 04:01 PM the voices made Joe Zitnik write:
>>>>>> I can't get here:
>>>>>> Is rulesemporium having issues again?
>>>>> I can rarely get there (via a browser). So rarely the site is almost
>>>>> useless.
>>>>> I've been having intermittent issues getting there from home for a
>>>>> while. Last time it happened, the site was down. I still can't get
>>>>> there....
>>>> Hum. I just tried again, and didn't have any problems this time
>>>> either.
>>>> Guess I'm lucky.
>>> Perhaps you are. I get "500 Server closed connection without sending
>>> any data back" or "500 Can't connect to
>>> (connect: timeout)" at least once an hour out of three queries an hour.

>> Daryl, I've tried before to tell you and other people RDJ is broken.

> Actually, you've not, and if you did it would be a waste of time given
> that I don't use RDJ and actually provide the sa-update channels for SARE
> rules.
>> Put
>> a 1 second sleep between each file fetch and see if that improves things.

> If you weren't in a hurry to make 3 posts about the same thing to the same
> thread, you'd see that I wrote that I'm seeing the timeout in 1 (or more)
> of ONLY THREE QUERIES AN **HOUR**. I've already got a 20 minute delay
> between queries. I'll try adding a 1 second delay to that though.
>> It keeps you from looking like a DoS attack.
>> Since I put that hack in my script has never failed me.

> As has been noted already, by Dallas, it's a problem with at least one of
> their network links being saturated by the DoS, not the DoS protection.

Is this perhaps a difference in wget and curl? This is an off hour. But I
am running again with no problems so far. (A friend put in the delay and
it worked for him, too.) Yeah, just finished faster than earlier today.

What was happening to me was nice fast progress through the first few of
my long list. Then it would start showing the timeouts for all the rest.
It was pure hunch that led to the delay strategy. And it has appeared to
work. I've never seen a timeout since then.

Go figure. It's magic? I dunno.