Hi Alex,

thanks for your interest.

You certainly have a point: The simplest way to do backups is an external
hard disk. However, this doesn't protect you from disaster like fire or
theft. Ok, the socks drawer as you proposed might give you some degree of
protection against burglary - at least as long as there's no suspicious
cabling ;)

You are right that the average user may not care about a disaster recovery
plan. Also, most households today don't have the upstream internet
bandwidth to conveniently upload terrabytes of data. So for the backup
use-case I'm thinking more of small businesses that also currently use the
NAS box backup strategy but should actually care for disaster protection.

These businesses often have better internet connections while the amount
of data to be transferred (with differential backup techniques) is not very
high. However, setting up an own remote storage infrastructure is usually
not economic and relying on cloud storage services introduces new trust
problems. So small businesses need simple-to-use, inexpensive and
trustworthy remote storage.

And I think a P2P file store does not have to be any more difficult to use
than a NAS box. When we envision an open-source software, NAS box
manufacturers could integrate the peer software directly into the box. So
setup of the system could be as easy as a traditional NAS box: You connect
power and network, then run the usual quick setup throughout which you
decide what fraction of the total storage to use locally and how much to
devote to the P2P system. You could then keep a local copy of your backups
as you're used to and benefit from quick response times and high data rates
while the system also transparently transfers your data into the P2P store
in the background for disaster protection.

But I agree: The user base for such backups may probably be limited.
However, I think a P2P global file store could also be the base for other
distributed computing services that could be built on top. The backup
use-case is just the most simple primitive and I've chosen it for
simplicity and because at least some other software is already available
in this area for comparison.

To go beyond the backup use-case, I think a P2P file store could be
interesting also for other mostly read-only data like for example media.
So I think there is merit in storing media like pictures, music and videos
just once and in a remote location. It may be nice to be able to connect
to your music collection easily from different devices like your mobile
phone, computer or home entertainment system instead of managing different
sub-collections on different local storage media. Also you might like to
have access from anywhere like when on travel or at a friend's house.

This is by no means a new idea and there are cloud storage services under
development or already available that provide you exactly this. However,
because such service requires rather much storage you won't get this for
free. Also, there's a trust problem as well even though it's "just" media
data:

If a central storage provider knows which media you store in his system he
will be able to deduce your interests (and use for advertising, for
instance). Further, the storage provider could be forced or feel compelled
to delete some of your data - for example because of alleged copyright
infringement. I'd just like to point to the Amazon / Kindle story when
Amazon deleted Orwell's "1984" from its customers e-readers. I believe
we'll see similar stories with cloud storage in the future. And finally,
there's also the risk of data suddenly becoming inaccessible because of
insolvency or bankruptcy that is inherent to all cloud storage services.

Therefore, I think a P2P file store may be well suited not only for
backups but also to store other immutable data like media. Especially for
media one could also leverage de-duplication effects (more than for backups)
in order to reduce the necessary upstream bandwidth and to save overall
storage requirements. Further, media is uploaded just once and then
infrequently read, so may be better suited to the upstream-limited internet
links of private households...

Best regards,
Michael


Quoting Alex Pankratov <[email protected]>:

>> -----Original Message-----
>> From: [email protected]
>> [mailto:[email protected]] On Behalf Of
>> Michael Militzer
>> Sent: January 26, 2011 11:04 AM
>> To: [email protected]
>> Subject: [p2p-hackers] P2P file storage systems
>>
>> Hi all,
>>
>> I am new to the list and also have no background with P2P. However,
>> I'd like to realize a P2P related project and have therefore read a
>> bit on the topic during the past year.
>
> Welcome to the list, Michael.
>
>> Basically, what I have in mind is a peer-to-peer, wide-area persistant
>> file storage system for a backup use-case.
>
> This is *very* interesting area from a technology perspective,
> but pragmatically speaking it has virtually no demand. When it
> comes to backing up few TBs of data the simplest solution is
> to buy and use an USB drive. If a physical security of the USB
> drive is an issue - get NAS box and WiFi router and stash them
> in the socks drawer :)
>
> In other words, the first question is if you could elaborate
> why you think the system like one you sketched out is needed
> (which in turn leads to what you exactly mean by 'untrusted').
>
> I am all for discussing finer technical points, just want to
> understand a larger picture first.
>
> Alex
>
> _______________________________________________
> p2p-hackers mailing list
> [email protected]
> http://lists.zooko.com/mailman/listinfo/p2p-hackers
>
>



_______________________________________________
p2p-hackers mailing list
[email protected]
http://lists.zooko.com/mailman/listinfo/p2p-hackers

Reply via email to