Errors @ http://fah-web.stanford.edu/cgi-bin/main.py?qtype=t

[BrokenWolf]

When I try to go to http://fah-web.stanford.edu/cgi-bin/mai ... =teamstats
http://fah-web.stanford.edu/cgi-bin/mai ... =userstats
http://fah-web.stanford.edu/cgi-bin/mai ... pe=osstats

I get the following message:

403 Forbidden
Forbidden
You don't have permission to access /cgi-bin/main.py on this server.(or whatever stats I am trying to look up)


I have tried this on systems on my Business DSL and Residential DSL connections (2 entirely separate connections) and different browsers on the systems as well.

Am I smoking something and not realizing it?

Broken

[bruce]

Perhaps a variation of this issue.
viewtopic.php?f=16&t=13472

If you've over-used the cgi-bin pages you probably have been banned, although I have never seen a 403 error associated with that issue before.

When I follow the links that you gave, I see a page saying that the Stats are being updated.

-------------

Or, somewhere I saw a problem report about a Cisco server setting that blocked access to Stanford.edu because it was a proxy server.

[John_Weatherman]

Works fine for me ... what's your way to connect and have you proxies, firewalls or the such like to mess things up?

[P5-133XL]

Actually with that forbidden code, II wonder if you were one of the people banned for using scripts? http://folding.typepad.com/news/2010/02 ... ripts.html

[BrokenWolf]

Well I do not use any scripts to get any stats. But the funny part is why would a dynamic IP address be banned as well unless a whole range has been banned? My static/business I could see but my Residential DSL has nothing going to F@H or looking up stats.

Broken

[sgb101]

why could you get banned for viewing http://fah-web.stanford.edu/cgi-bin/mai ... ame=sgb101page to many times.

how many is to many? ive been checking their about 4 times a day for the as long as i can remember.

[BrokenWolf]

No proxies nor has anything changed on either DSL line.
Well I have not been checking the stats any more than normal. The main thing is that not only did my business DSL get blocked if that is the case but so did my residential DSL, which is using an entirely different IP range and is not used for any Folding @ home activities and is only used by my sons for their internet usage and online gaming.

Now how do I go about getting my static IP and my home dsl IP range unblocked?

[bruce]

why could you get banned for viewing http://fah-web.stanford.edu/cgi-bin/mai ... ame=sgb101page to many times.

how many is to many? ive been checking their about 4 times a day for the as long as i can remember.

The general answer has been that the cgi pages should be used manually, not by script. Four times a day is fine.

Those who run 3rd party stats sites want stats for EVERYBODY and they wanted it MANY TIMES per day. Their script can download the flat files which are generated whenever the stats are updated. The 3rd party sites (and anyone, for that matter) use local computer resources to parse them and generate any information that you find interesting but they can't submit thousands/millions of individual queries to Stanford's database.

I'm not sure if there is a specific number that distinguishes a script from a manual query. The stats can bog down from too much load. In effect, a script can create something like a DoS attack, and the server needs to be protected from that sort of activity. The database is a shared resource with enough capacity to handle "normal" accesses, but that doesn't mean that it should be sized to handle any possible combination of accesses that might be created by scripts.

[BrokenWolf]

Well again, no scripts are in use on any of my systems to pull stats. I only use a web browser and I had been at work for 12 hours yesterday and when I got home to check that is what I got.

Now, how to get this fixed/unbanned?

[7im]

Are you using any widgets or client monitoring tools that might be trying to download project info, or points, that kind of thing? You may not be accessing the CGI pages, but a tool you are using might be...

[BrokenWolf]

The only thing I have been using is HFM.net for monitoring my systems. I have since closed the app and asked for an update to see if it is still requesting the WU info. Below is what is happening. I am at work right now so I can only do limited troubleshooting.

<my static IP>[17/Mar/2010:09:41:22 -0800] "GET /cgi-bin/fahproject?p=6012 HTTP/1.1" 403 228
<my static IP>[17/Mar/2010:09:41:22 -0800] "GET /cgi-bin/fahproject?p=6014 HTTP/1.1" 403 228
being repeated and repeated.

BW

[bruce]

I suspect that there may be a problem associated with HFM if you're generating web pages with it, but I don't use that feature and I have not done a lot of research into that possibility. Does that apply to you?

EDIT: My question was already answered. How frequently were you regenerating your web pages? (The default was 15 minutes.)
We'll see what happens next.

[BrokenWolf]

Ah Ha. That is what it was folks. Some time back I had set it to generate a web page every 3 hours (120 minutes) to monitor my client while I was in the office. Turned it off a month or 2 ago (or so I thought). I started HFM back up and looked and the configuration. It is (was) set to 1 minute so instead of turning it off I fat fingered it.

I'd like a little ketchup with my crow please.

[7im]

It would be nice if the HFM developer would set the minimum value to that feature to 1 hour to avoid pounding the Stanford servers, on purpose, or by mistake.

[VijayPande]

Actually, the HFM developer (and all developers) should NEVER query any of our cgi-bin pages. All it takes is some bug on their side to wreak havoc. All developers should honor the robots.txt we have on our web site.

Sorry to be so hard on this, but this creates huge problems for our donors and major distractions for my team, who should be doing science, not dealing with problems caused by 3rd party apps.