Rebecca -> good analogy, however, I think this is conceptually different.
This is more like the standing there "guarding", while at the same time
pondering his metaphysical existence. (Or more likely what he's going to do
when he gets home, or why that bee is flying that way instead of this way
... you get the idea.) Obviously computers are much better "thinkers" in
this respect since they would immediately handle an "interrupt" much better
than a zoned-out guard.
I have thought of doing this as well, with protein-folding or SETI (who
knows? There might be life out there ;). Especially since my box is a
dually P-II (only one processor is getting used right now). MY biggest
problem has been getting an SMP-enabled kernel on there, but that probably
doesn't apply to you.
The biggest security hole is going to be when your IPCop goes out for
updates, or to "drop-off" the numbers it has crunched. But I don't think
that there are any related vulnerabilities with this specifically (at least
none that I have heard of). On the first page of the WebUI you can get the
load average information (order is <min><max><avg>). If yours is anything
like mine, it's probably in the 0.2-0.6 percent range. And you certainly
have plenty of processor overhead to do something like this (if you want).
I know that the Linux version of SETI is a single executable, and I believe
the same applies for Folding. So I would scp the executable over, be sure
to set the permissions (chmod a+x folding.exe) and run it as a background
process (nohup folding.exe &).
After doing this though, I would watch the processor usage on the first page
of the WebUI (it should go up to about 100%), snort (assuming you have it
implemented), and maybe take a few tcpdump captures to see if you're
dropping any packets. Other than that, you should be good to go.
-= tH3 D@...; =-
[mailto:ipcop-user-admin@...] On Behalf Of Rebecca A. Ward
Sent: Saturday, August 02, 2003 12:47 PM
Subject: Re: [IPCop-user] Protein Folding
Charlie King <charlie@...> wrote:
> It strikes me that my IPCop gateway is probably spending lots
> of its CPU cycles twiddling its digital thumbs...
> Perhaps that time could be better spend twiddling with digital
> proteins (http://folding.stanford.edu/). Briefly, this is a
> distributed computing exercise along the same lines as
> SETI@..., but with rather more terrestrial goals - to
> accurately simulate the behaviour of proteins, and thereby
> advance medical understanding into diseases as varied as
> Alzheimer's, Mad Cow (BSE), CJD, ALS, and Parkinson's.
> I'm not clued up enough, however, about the workings of IPCop
> to understand a) if there would be security implications in
> running the client (I'm just a home site, looking to avoid
> being a DDoS client) and b) how to go about installing the
> linux client.
> Any comments from the clued-up?
While I would never presume to tell someone what they can or
cannot do with their own IPCop box, I will offer this bit of
If you hire a guard to protect your front door to make sure no
one comes in who isn't worthy. He/she will spend a lot of time
just standing there. But if you use his time to go and fetch
slippers, fold your laundry, or do other tasks, while he will no
longer be idle, he also will not have his eye on your front door
100% of the time. So, if someone sneaks in, don't blame him.
This SF.Net email sponsored by: Free pre-built ASP.NET sites including
Data Reports, E-commerce, Portals, and Forums are available now.
Download today and enter to win an XBOX or Visual Studio .NET.
IPCop-user mailing list