[OpenDHT-Users] OpenDHT storage question

Ahrenholz, Jeffrey M jeffrey.m.ahrenholz at boeing.com
Fri Dec 14 19:27:53 UTC 2007


Sean,
Thanks for the response. We'll munge these values before using them as
keys.

I'm assuming the server pads the key value out to 160 bits when it
receives a shorter key?

-Jeff

> -----Original Message-----
> From: Sean Rhea [mailto:sean.c.rhea at gmail.com] 
> Sent: Friday, December 14, 2007 11:15 AM
> To: Ahrenholz, Jeffrey M
> Cc: opendht-users at opendht.org
> Subject: Re: [OpenDHT-Users] OpenDHT storage question
> 
> Jeff,
> 
> Yes, key-value pairs are partitioned across nodes by their key's
> prefixes.  If you want to balance the load from HIP-style keys (and
> you should), you need to reverse them and pad the tail up to 160 bits
> total.  I.e.:
> 
> Key = HASH | PREFIX | 32 bits of zeros
> 
> After that, it should work fine.
> 
> Cool application idea, by the way.
> 
> Sean
> 
> 
> 
> On Dec 14, 2007 10:33 AM, Ahrenholz, Jeffrey M
> <jeffrey.m.ahrenholz at boeing.com> wrote:
> > Hi, I've been working with the HIP protocol and have worked out some
> > extensions to store some HIP information (basically the 
> hash of a public
> > key and an address) into OpenDHT. This enables one type of 
> resolution
> > service for HIP...
> >
> > My question is, if I am storing several DHT records using a key that
> > looks like a fixed 28-bit prefix concatenated with 100-bits 
> of random
> > number, will the fixed prefix cause records to be stored on the same
> > server?
> >
> > (The prefix will always be the same, coming from RFC 4843; 
> key = 128-bit
> > HIT = (PREFIX | HASH) where "|" indicates concatenation)
> >
> > Apologies for my lack of understanding of the Bamboo DHT 
> software, and
> > if this is covered by a publication or on this mailing list :)
> >
> > thanks,
> > -Jeff
> > _______________________________________________
> > OpenDHT-Users mailing list
> > OpenDHT-Users at opendht.org
> > http://opendht.org/cgi-bin/mailman/listinfo/opendht-users
> >
> 
> 
> 
> -- 
> "Humanity has advanced, when it has advanced, not because it has been
> sober, responsible, and cautious, but because it has been playful,
> rebellious, and immature." -- Tom Robbins
> 


More information about the OpenDHT-Users mailing list