Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Show HN: This website is hosted on DNS

banner.triweb.dev

Post reply on HN
Print view View original

11–20 of 28 posts

Re: Show HN: This website is hosted on DNS

#11

Post by quickthrower2 »

plagiat0r wrote: post #10
plagiat0r wrote: post #9

Earlier quoted context omitted.

I should elaborate more about why I believe this is a bad idea to use DNS as storage / content. First, because of a distributed nature of the DNS, clients do not directly fetch this data. Instead, resolvers and DNS forwarders do the work and cache it. Secondly, because displaying or converting this data to html requires a special website or a dedicated client, at which point this dedicated component should use a diff…

Let me just add one more thing why this idea seems bad to me personally. I'm not atacking you, but the idea. So the way things work today is this, the way I see it: 1. We have a physical layer -> a wire 2. We have a data link layer -> ethernet + ARP or NDP are common here 3. we have a network layer -> currently, two networks are widely used: IPv4 and IPv6 4. We have a transport layer: tcp and udp and quic 7. We have…

Thanks that is a cool insight about what layer DNS should be thought of as. Never even thought about this.

Re: Show HN: This website is hosted on DNS

#12

Post by plagiat0r »

quickthrower2 wrote: post #11
plagiat0r wrote: post #10

Earlier quoted context omitted.

Let me just add one more thing why this idea seems bad to me personally. I'm not atacking you, but the idea. So the way things work today is this, the way I see it: 1. We have a physical layer -> a wire 2. We have a data link layer -> ethernet + ARP or NDP are common here 3. we have a network layer -> currently, two networks are widely used: IPv4 and IPv6 4. We have a transport layer: tcp and udp and quic 7. We have…

Thanks that is a cool insight about what layer DNS should be thought of as. Never even thought about this.

I really do not want to start a flame war about OSI model, but I tent to bend the OSI model and interpret the "layers" by the function they presents.

Another example, lets say you spin up a QuicTun, nacl-based encapsulation tunel, encrypted, running over IP + UDP as a layer7 process, in userland, there is no kernel implementation. For me, this vpn is Layer2 (when tap mode) or Layer3 (when tun mode) regardless that it uses all 7 layers to establish the tunnel. Sure it uses all layers to establish the connection but once it is done, I have layer3 device for further use.

Same for DNS - for me it is Layer3 because of the crucial function it provides, so I could rather say it is functionally on the same layer as IP in a sense, it is crucial to make a connection. Does not matter that it uses all 7 layers to get me the network (A)ddress / (AAAA)ddress, it is still functionally a network layer for me. Only once this is establised, I can exchange data.

And while DNS is running on layer7 - for me personally, the function it performs is "to allow connect() to a string as opposed to a network address (IP/IPv6)" and because of that, it is a Layer3 for me on the functional level, same as a VPN software to access some corporate stuff. Does not matter that VPN uses TLS and DNS under the hood and span across all 7 layers to encapsulate my data, functionally it is layer3 for me.

I do a lot of corporate DNS trainings and people are surprised about this view. Software developers are unhappy that we start the "DNS course" with IP basics, unicast, anycast, 101 overview of BGP, ARIN, provider independed vs Provider Aggregatable address space. Developers would rather learn about DNS, not what an IP address is. Network guys sits happy on the course, but they are not happy to hear my view that:

"there is no DNS without network and there is no network without a DNS"

However, slowly but surely once the full training is finished, both sides understand that IP/IPv6 and DNS are in a symbiosis and can't be separated. We touch a lot of topics, including lack of interoperatibilty between IPv4 and IPV6, DNS64, XLAT464, etc.

Sure, if you run an air-gapped network in some military facility, you can get things working without DNS, but 99.999% networks can't work without a DNS.

Re: Show HN: This website is hosted on DNS

#13

Post by coretx »

zzem wrote: post #6
Ringz wrote: post #2

Looks interesting! What are the advantages (besides the mentioned ones) and what are the disadvantages?

I think the main advantage is the ease and speed of use for people who want a simple, business-card-like website or a holding/landing page. The main disadvantage is that due to the UI/UX of domain control panels, managing anything more than a few paragraphs of text gets messy really quickly. But that could be actually an advantage, as the DNS is not particularly well suited to serve large amounts of data, and the app…

Let's be honest here. The main advantage is to make use of someone else his resources. Since it's potentially degrading critical infrastructure and increasing the cost of market entry: Unethical.

Re: Show HN: This website is hosted on DNS

#14

Post by whodev »

plagiat0r wrote: post #12
quickthrower2 wrote: post #11

Earlier quoted context omitted.

Thanks that is a cool insight about what layer DNS should be thought of as. Never even thought about this.

I really do not want to start a flame war about OSI model, but I tent to bend the OSI model and interpret the "layers" by the function they presents. Another example, lets say you spin up a QuicTun, nacl-based encapsulation tunel, encrypted, running over IP + UDP as a layer7 process, in userland, there is no kernel implementation. For me, this vpn is Layer2 (when tap mode) or Layer3 (when tun mode) regardless that it…

> Sure, if you run an air-gapped network in some military facility, you can get things working without DNS, but 99.999% networks can't work without a DNS.

It's funny you say that, because in the air-gapped network in some military facility I work in, we actually replicate certain public DNS entries to have some specific systems and services working for our developers.

Re: Show HN: This website is hosted on DNS

#15

Post by plagiat0r »

whodev wrote: post #14
plagiat0r wrote: post #12

Earlier quoted context omitted.

I really do not want to start a flame war about OSI model, but I tent to bend the OSI model and interpret the "layers" by the function they presents. Another example, lets say you spin up a QuicTun, nacl-based encapsulation tunel, encrypted, running over IP + UDP as a layer7 process, in userland, there is no kernel implementation. For me, this vpn is Layer2 (when tap mode) or Layer3 (when tun mode) regardless that it…

> Sure, if you run an air-gapped network in some military facility, you can get things working without DNS, but 99.999% networks can't work without a DNS. It's funny you say that, because in the air-gapped network in some military facility I work in, we actually replicate certain public DNS entries to have some specific systems and services working for our developers.

Of course, you want SNMP monitors get you and email as an alert, you have to setup auth DNS with xyz.local and use it in air-gapped environments for your alert mails. Hard drive failure?, RX/TX errors on a switch port? fire up mail over SMTP with an SNMP agent, etc. the list goes on and on.

Network without DNS is un-useable. Apologies for my english, it is not my native language.

Re: Show HN: This website is hosted on DNS

#16

Post by whodev »

plagiat0r wrote: post #15
whodev wrote: post #14

Earlier quoted context omitted.

> Sure, if you run an air-gapped network in some military facility, you can get things working without DNS, but 99.999% networks can't work without a DNS. It's funny you say that, because in the air-gapped network in some military facility I work in, we actually replicate certain public DNS entries to have some specific systems and services working for our developers.

Of course, you want SNMP monitors get you and email as an alert, you have to setup auth DNS with xyz.local and use it in air-gapped environments for your alert mails. Hard drive failure?, RX/TX errors on a switch port? fire up mail over SMTP with an SNMP agent, etc. the list goes on and on. Network without DNS is un-useable. Apologies for my english, it is not my native language.

> you have to setup auth DNS with xyz.local and use it in air-gapped environments for your alert mails

True, but I was talking more about creating DNS entries for NPM (and other package managers) and redirecting them to our internal services. I just thought it was funny to mention since we impersonate public sites on our internal air-gapped network.

Re: Show HN: This website is hosted on DNS

#17

Post by Kharacternyk »

zzem wrote: post #5
Kharacternyk wrote: post #4

This is a nice PoC. For how long are TXT records usually cached? Triweb might not see the latest changes to the websites for quite some time.

Thanks! The time for which TXT records are cached is determined by their TTLs. When publishing a DNS TXT record, you can usually set the TTL to be as low as 1 minute or even less, so any changes to the content would be picked up really quickly. ( https://developers.cloudflare.com/dns/manage-dns-records/ref... )

I know that it's all configurable, what I meant is that some DNS providers may apply a long TTL by default, and it may be desirable to let users of Triweb know about this possible source of confusion.

Re: Show HN: This website is hosted on DNS

#19

Post by zzem »

plagiat0r wrote: post #10
plagiat0r wrote: post #9

Earlier quoted context omitted.

I should elaborate more about why I believe this is a bad idea to use DNS as storage / content. First, because of a distributed nature of the DNS, clients do not directly fetch this data. Instead, resolvers and DNS forwarders do the work and cache it. Secondly, because displaying or converting this data to html requires a special website or a dedicated client, at which point this dedicated component should use a diff…

Let me just add one more thing why this idea seems bad to me personally. I'm not atacking you, but the idea. So the way things work today is this, the way I see it: 1. We have a physical layer -> a wire 2. We have a data link layer -> ethernet + ARP or NDP are common here 3. we have a network layer -> currently, two networks are widely used: IPv4 and IPv6 4. We have a transport layer: tcp and udp and quic 7. We have…

Thank you for your insights. I understand the concerns you've raised, but there's also another side to DNS that's often overlooked: its role as a globally available, distributed database for storing small amounts of data linked to domain names.

Now, I'm not saying that people should start to host their files on DNS, but for simple, personal "banner" like pages that say "Hi, I'm X, love cats, catch me on @email" that should be OK, may be easier than setting up a separate hosting, and make the web more inclusive for non-technical users. Storing and transferring a few hundred bytes for this purpose under a distinct namespace/zone in a niche use case should be well within the capacity of the existing DNS software and infrastructure. We already have other standardized, dedicated classes of informational ("Layer 7 data") DNS records like HINFO and RP, so why not store a piece of text that would have a meaning for humans and present it in a nicely formatted (themes) and accessible (HTTP) way.

The way I see it is as a scale: DNS TXT records may be ok to use for small, simple, informational pages; - static hosting (GitHub Pages, Cloudflare Pages, etc.) should be used for longer texts, pages with custom assets (e.g., images), or multi-page sites; - SSGs and dedicated platforms for blog and bigger websites (Jekyll, Wordpress, etc.); - VPSes, EC2, Lambda, Firebase etc. for bigger things; - ending probably with dedicated servers, networking, and own ASNs for big online platforms. There is always some amount of data or activity above which one should switch to the next level, but switching to it requires additional knowledge and resources; and the other way around, buying and maintaining a dedicated server just to host a simple, static website with low traffic is an overkill.

Also, one thing about the app's architecture may be worth highlighting: The app uses a pre-set DoH provider (currently Cloudflare) to resolve DNS queries on the client-side, so the impact on the DNS ecosystem is close to none as only the source DNS server and the DoH resolver store and transfer these TXT records. Such setup also protects sites against mitm attacks. If the traffic becomes too big for Cloudflare infrastructure (if that's even possible), we can always setup and switch to our own dedicated DoH endpoint. Similarly, if the DNS hosting provider thinks it's abusing their servers, they can increase pricing or limit the number and/or length of TXT records their customers can setup for a single domain.

Re: Show HN: This website is hosted on DNS

#20

Post by zzem »

XCSme wrote: post #18

I don't understand how this is "decentralized", the first step is "Point your domain at a triweb relay server". Doesn't this make the triweb relay server the central authority?

At the moment, it does, but we plan to release a Docker image of the relay server so that anyone would be able to spin his/her own relay or use any other existing public relay.

The triweb platform is still a work-in-progress, and there is a lot left to be done, released, and documented. The Banner app is mostly just an early demo of how TWAs are built and how they may be deployed to domains, that I thought may be interesting to HN.

Post reply on HN