#1 2011-03-27 14:47:52

XVilka
Reverse Engineer
From: Moscow
Registered: 2010-10-03
Posts: 6
Website

SHA1 Collision search for RSA keys

Offline

#2 2011-03-28 09:51:28

poseidon
New member
Registered: 2011-03-28
Posts: 3

Re: SHA1 Collision search for RSA keys

you'll need to think about how to create work units, because creating all possible work units at once would take up lots of database space. Would be probably best to create work units in a defined order (e.g. increasing/decreasing) and on-demand (so that each time a client connects it either gets an old work unit (a replicated one) or a new one is created and sent to the client). There should be some kind of decision heuristic for deciding whether a client gets a replicated or a new work unit.

Furthermore there must be a timeout for work units because a client that connects, receives a work unit, but never connects again would otherwise "eat up" a work unit.

There are many things to consider, and I'm not sure if it wouldn't be easier to use the boinc infrastructure. other than needing a really fast SQL server, it's quite easy to manage and use.

Last edited by poseidon (2011-03-28 09:52:09)

Offline

#3 2011-03-28 10:08:11

mib_force
New member
Registered: 2011-03-28
Posts: 3

Re: SHA1 Collision search for RSA keys

Xvilkas ideas sound great, but i agree with poseidon. But i guess it would be to complex to implement a server like the BOINC infrastructure provides it. At least when it comes to managing the work units.

I also would prefer the BOINC server/client structure if possible

Offline

#4 2011-03-28 10:29:32

p0lse
New member
Registered: 2011-03-28
Posts: 2

Re: SHA1 Collision search for RSA keys

The problem with BOINC is its competitive culture. By not ranking users by pointless points there's less need for protecting the server against cheaters.

Offline

#5 2011-03-28 10:32:38

poseidon
New member
Registered: 2011-03-28
Posts: 3

Re: SHA1 Collision search for RSA keys

but the competitive nature makes users run the project for a longer time. Boinc has a pretty good anti-cheater system built in, the only thing you need to make sure is no one tampers with your worker. (I'm not sure if there are built in boinc functions for that, but it could be as easy as signing the result file and verifying it on the server)

Last edited by poseidon (2011-03-28 10:33:10)

Offline

#6 2011-03-28 11:13:16

p0lse
New member
Registered: 2011-03-28
Posts: 2

Re: SHA1 Collision search for RSA keys

By protecting the worker the user can no longer compile it for her esoteric platform of choice. That leaves the project admin to make builds for all combinations of OS and CPU arch. This usually means turning away anyone who doesn't run Win or Linux on x86(_64)...

Offline

#7 2011-03-28 13:01:21

mib_force
New member
Registered: 2011-03-28
Posts: 3

Re: SHA1 Collision search for RSA keys

Offline

#8 2011-03-28 17:47:28

XVilka
Reverse Engineer
From: Moscow
Registered: 2010-10-03
Posts: 6
Website

Re: SHA1 Collision search for RSA keys

Just received answer from BOINC team - server have no support any other DB engine, than MySQL sad

Offline

#9 2011-03-29 08:03:57

mib_force
New member
Registered: 2011-03-28
Posts: 3

Re: SHA1 Collision search for RSA keys

Offline

#10 2012-10-12 01:53:32

jorgea
New member
Registered: 2012-10-12
Posts: 1
MacintoshChrome 21.0.1180.90

Re: SHA1 Collision search for RSA keys

Offline

Registered users online in this topic: 0, guests: 1

Board footer

Powered by FluxBB
Modified by Visman

[ Generated in 0.018 seconds, 10 queries executed - Memory usage: 461.7 KiB (Peak: 463.93 KiB) ]