Re: [bitcoin-dev] BIP159 - NODE_NETWORK_LIMITED service bits, extendability

[Peter Todd via bitcoin-dev]

On Tue, Nov 21, 2017 at 06:45:33PM +, Gregory Maxwell via bitcoin-dev wrote:
> With the way pruning works today my expirence is that virtually no one
> sets any parameter other than the minimum, though even with that set a
> few more blocks can be available.

FWIW, I run all my pruned nodes with the prune parameter set to about a month
worth of blocks (a few GB). And come to think of it, I should bump that up even
higher now that segwit has increased the blocksize.

-- 
https://petertodd.org 'peter'[:-1]@petertodd.org


signature.asc
Description: Digital signature
___
bitcoin-dev mailing list
bitcoin-dev@lists.linuxfoundation.org
https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev

Re: [bitcoin-dev] BIP159 - NODE_NETWORK_LIMITED service bits, extendability

[Jonas Schnelli via bitcoin-dev]

Hi Sjors

Thanks for picking this up.

There where some previous discussions about this [1] [2].
Initially, the idea was to have two service bits to signal (up to three) states.
But, since it is not clear what use-cases the bits signalling >288 blocks would 
fulfil, I have limited BIP159 to a single 288blocks-available signalling.

Therefore, BIP159 aims to improve the block relay state around the tip (24h) 
which seems to be the most significant request peak (peers out of IBD).
Also, it takes an acceptable transition for pruned node operators into account. 
Once BIP159 gets active on the network, pruned peer operators may see an 
increase in CPU and bandwidth usage.

SPV peers may also connect to BIP159 nodes, scan the mempool and wait for 
unconfirmed transactions (they don’t do this now because pruned nodes don't 
signal any service).

Future extensions are possible. Maybe a p2p command that could tell more infos 
about the pruning state would be useful.

BIP159 also recommends to fix the fingerprinting weakness by fix limiting it to 
288 blocks, making it impossible for an attacker to fingerprint your peer by 
scanning how deep the peer can serve blocks. This may be a reduction for 
possible use cases with todays pruned peers and an idea would be to relax this 
limit for whitelisted peers (or peers connecting via BIP150 [not implemented], 
and this is the only connection between BIP150 and BIP159).

However, I think the scope of BIP159 should be kept as it is. More flexibility 
can be added later when we have gathered more information during BIP159 
deployment.
Also, the implementations is an advanced stage [3][4]

—


[1] https://botbot.me/freenode/bitcoin-core-dev/2017-04-27/?msg=84827228=3
[2] 
https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2017-May/thread.html#14314
[3] https://github.com/bitcoin/bitcoin/pull/10387
[4] https://github.com/bitcoin/bitcoin/pull/11740


> Am 21.11.2017 um 04:03 schrieb Sjors Provoost via bitcoin-dev 
> :
> 
> I came across the proposed Bitcoin Core implementation of BIP159 [0] in this 
> PR [1]. The goal is to allow pruned nodes to "serve a limited number of 
> historical blocks" (as opposed to none at all).
> 
> It contains a counter-measure for peer fingerprinting. I'm trying to 
> understand how that impacts extendibility.
> 
>> Peers may have different prune depths (depending on the peers configuration,
>> disk space, etc.) which can result in a fingerprinting weakness (finding the
>> prune depth through getdata requests). NODE_NETWORK_LIMITED
>> supporting peers SHOULD avoid leaking the prune depth and therefore
>> not serve blocks deeper then the signaled NODE_NETWORK_LIMITED
>> thresholds.
> 
> This means pruned nodes can only serve the last 288 blocks:
> 
>> If signaled, the peer MUST be capable of serving at least the last 288 
>> blocks (~2 day
> 
> As the blockchain keeps growing there will be ever more pruned nodes (perhaps 
> offset by new nodes with more storage).  Although a strict improvement over 
> todays situation, it seems a bit wasteful to have a node with 10-100 GB of 
> storage only be able to share the most recent 288 blocks.
> 
> It would be nice if a future extension of this BIP allows more flexibility. 
> To limit the ability to fingerprint nodes, we could limit the number of 
> choices to e.g. 288 + 1000 * 2^n. That yields only 8 possibilities at the 
> current chain size. A slightly better formula could take into account typical 
> hard drive size increments, leaving enough space for the OS and other data. 
> Node operators could opt-in to this if they think the increased fingerprint 
> risk outweighs their desire to share archived blocks.
> 
> I can also imagine - but not implement :-) - a future scenario where nodes 
> prune a random subset of their chain, meaning that even nodes with little 
> storage can be of help during Initial Blockchain Download (IBD) of other 
> nodes.
> 
> 
> How would such extension be signaled for? Would we need a whole new version 
> bit?
> 
> Would upgraded nodes need a new message type to communicate the chosen prune 
> depth? Or can that information tag along some existing message?
> 
> Jonas Schnelli pointed out on the Github discussion that waiting for BIP150 
> would be appropriate. Can you explain how this is related? Although I can see 
> why whitelisted peers can be exempted from the anti-fingerprinting measure, I 
> would not want to restrict it to just those.
> 
> 
> Some minor suggestions for improving the BIP itself:
> * add link to mailinglist discussion(s) in reference section
> * explain that 288 is not just the minimum limit for Bitcoin Core, but also 
> the bulk of traffic (as I understand from earlier discussion [2])
> 
> Cheers,
> 
> Sjors
> 
> [0] https://github.com/bitcoin/bips/blob/master/bip-0159.mediawiki
> [1] https://github.com/bitcoin/bitcoin/pull/10387
> [2] 
> 

[bitcoin-dev] BIP159 - NODE_NETWORK_LIMITED service bits, extendability

[Sjors Provoost via bitcoin-dev]

I came across the proposed Bitcoin Core implementation of BIP159 [0] in this PR 
[1]. The goal is to allow pruned nodes to "serve a limited number of historical 
blocks" (as opposed to none at all).

It contains a counter-measure for peer fingerprinting. I'm trying to understand 
how that impacts extendibility.

> Peers may have different prune depths (depending on the peers configuration,
> disk space, etc.) which can result in a fingerprinting weakness (finding the
> prune depth through getdata requests). NODE_NETWORK_LIMITED
> supporting peers SHOULD avoid leaking the prune depth and therefore
> not serve blocks deeper then the signaled NODE_NETWORK_LIMITED
> thresholds.

This means pruned nodes can only serve the last 288 blocks:

> If signaled, the peer MUST be capable of serving at least the last 288 blocks 
> (~2 day

As the blockchain keeps growing there will be ever more pruned nodes (perhaps 
offset by new nodes with more storage).  Although a strict improvement over 
todays situation, it seems a bit wasteful to have a node with 10-100 GB of 
storage only be able to share the most recent 288 blocks.

It would be nice if a future extension of this BIP allows more flexibility. To 
limit the ability to fingerprint nodes, we could limit the number of choices to 
e.g. 288 + 1000 * 2^n. That yields only 8 possibilities at the current chain 
size. A slightly better formula could take into account typical hard drive size 
increments, leaving enough space for the OS and other data. Node operators 
could opt-in to this if they think the increased fingerprint risk outweighs 
their desire to share archived blocks.

I can also imagine - but not implement :-) - a future scenario where nodes 
prune a random subset of their chain, meaning that even nodes with little 
storage can be of help during Initial Blockchain Download (IBD) of other nodes.


How would such extension be signaled for? Would we need a whole new version bit?

Would upgraded nodes need a new message type to communicate the chosen prune 
depth? Or can that information tag along some existing message?

Jonas Schnelli pointed out on the Github discussion that waiting for BIP150 
would be appropriate. Can you explain how this is related? Although I can see 
why whitelisted peers can be exempted from the anti-fingerprinting measure, I 
would not want to restrict it to just those.


Some minor suggestions for improving the BIP itself:
* add link to mailinglist discussion(s) in reference section
* explain that 288 is not just the minimum limit for Bitcoin Core, but also the 
bulk of traffic (as I understand from earlier discussion [2])

Cheers,

Sjors

[0] https://github.com/bitcoin/bips/blob/master/bip-0159.mediawiki
[1] https://github.com/bitcoin/bitcoin/pull/10387
[2] 
https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2017-May/thread.html#14315


signature.asc
Description: Message signed with OpenPGP
___
bitcoin-dev mailing list
bitcoin-dev@lists.linuxfoundation.org
https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev