I like to host Hive seed nodes :) It's a small thing that I do for Hive to help out the platform's health and keep the engine running. Seed nodes are crucial to the Hive ecosystem and are what other nodes sync from. By hosting multiple of these, I'm ensuring that even if one goes down, there's others that can be used by the Hive nodes to get the block data to keep running.
On one of older posts I mentioned that I was using witness votes to try and bring back the NYC seed node. Well, NYC colocation is almost ready but I don't have the funds to buy a server so I'm hoping that the DHF will help pay for the hardware.
I do want to iterate that this DHF proposal is just for the hardware itself. I will be covering the costs of running the server for it's lifetime myself(though witness votes to help cover the cost are greatly appreciated).
The hardware that will be purchased is a Dell server with a Xeon Gold 6143 processor. I've gone with Ryzens for the rest, but going to be trying Intel on this one. Also will be using NVME drives. The total cost shown by the site is $2,109. With taxes, crypto conversion fees and remote hands installation fees, I'm asking for $3,000 to cover the whole machine. For time reasons, I'd like to get this purchased and ideally delivered before the second week of September so I'm asking for it as $600/day for 5 days, with the proposal pay starting from Monday August 18th.
My current seed nodes are in San Fransisco and Chicago(I also have 2 more running in Stockholm and Amsterdam but they aren't running on owned hardware and so aren't part of my "cluster" officially, just there to be backups if everything breaks). With this in NYC, we should have pretty decent coverage in the US.
If you agree with this, please support proposal 354, voting links attached:
Peakd: https://peakd.com/me/proposals/354
Hive.Blog: https://wallet.hive.blog/proposals
Ecency: https://ecency.com/proposals/354
I've been advocating for infrastructure since day one. However, I'm not a fan of doing it this way. Seed nodes should be a shared responsibility of Hive witnesses. Not only those in top20, but also aspiring ones from top50 or so. It's great that you provide one, and in fact it's one of the reasons I'm voting for you.
But this proposal, rubs me the wrong way.
First about my PoV on seed nodes: you are running more than one public seed node. That's great, but running more of them doesn't really help the network that much. Sure, handful of healthy seed nodes are needed, and we should have few more, especially those that can act as "initial, default seed nodes". However, these should ideally be maintained by different independent entities. Otherwise such a redundancy serves little to no purpose for the platform, and doesn't meaningfully contribute to decentralization efforts.
(instead of funding this proposal, I encourage users to review their witness votes, and ask whether their witnesses provide public seed nodes. For example, @rishi556 does).
By the way, every hived connected to the network serves as p2p node and in many cases as a seed node, just not as mentioned "default, initial" one (for that a static address is required, preferably with own FQDN). I run plenty of these myself, and I know a few other witnesses / developers do as well. Thanks to the great core development efforts over time (kudos to @blocktrades and @thebeedevs teams), the hardware requirements for running consensus nodes (that is seed nodes, witness nodes, etc) have been drastically lowered, despite continuous growth of our chain. These improvements allow us to keep our hardware specs reasonable.
And reasonable is what this proposal lacks. The server itself is way, way, way overscaled for the claimed task. Seriously, it has at least 8x more CPU cores, 16x more RAM, and 2x more storage than a Hive seed node actually needs. I don't even use such powerful hardware for my public API node, not to mention my witness nodes.
I've featured my Hive consensus node setup that consumes 7W of power quite some time ago, and others have also written about efficiency, such as @themarkymark's take, and @borislavzlatanov's follow-up. Whenever there are witnesses, core developers, application builders, or even power users, then our p2p network will grow bigger and stronger organically, just by having them around.
Paying for such an overpowered server from DHF, one so powerful it wouldn't even notice a Hive seed node running on it, just isn't right.
I fully agree on having them run by multiple entities and try to encourage others to run nodes as well. If anyone else wants help running one please ask, I will help you out.
As for reasonableness, this was the best I can see for putting in a data center. I know it looks overspec'd at first glance. I thought the same too, but it's the best I can find within a "reasonable" budget. I need something with remote management and something that's reliable long term. From past experience, I've found that spending a bit more to get something reliable is worth it for colocating hardware, since the labor costs add up quick if you can't go in to do the work yourself(and sadly I live a flight away from NYC to be able to visit for fixing issues). Dells have been the most reliable hardware I've used so I went with them. Got plenty of broken gear from other brands, but none of my Dell servers have given me problems yet(knock on wood 🤞). The CPU was chosen for it's higher frequency compared to the other options. Sure it has more cores, but I've noticed faster sync times with faster frequency and I plan on keeping this around for as long as it's reasonable so spending a extra few bucks now for any future needs seems worthwhile to me. Upgrading in the future will cost more than getting higher performance now. I've also had enough used RAM fail on me that I've decided to overspec and trash out falty sticks. It's a lot easier & cheaper to ask the DC to take out half the sticks and see if it boots back up and swap the half taken out if it doesn't than working with them to narrow down which is the faulty stick. Same thing on the drives, while it has more than necessary, it's again cheaper long term to just disable the drives that die than replacing. When work is $300/hour with 1 hour minimum, I'd like to be able to fix things without needing to involve DC techs. As for why I keep my nodes in datacenters, my home network has a upload cap of 20mbps and I'd like to keep as much of that for my usage than machines.
Long term, this node will be up there, just later if this proposal doesn't get funded. I use Hive enough that I want it running and I'll play my part on keeping it running.
This might be a daft question, because I'm more a creative type than a technical one, but why does the seed node have to be in NYC if it's a flight away ? As long as it's one of several scattered across the globe, why can't it be at your place, or at least in the same town so that you can do any maintenance work yourself (or at least get it done locally) ?
Generally those nodes are best to be placed near those who need them. That's pretty much I mean writing about organic growth of the network. Developer needs them - developer runs them, etc.
I have a few. I’m getting colocation in New York for other reasons and I was planning on putting up a Hive Seed node here to help the network out. I live in Chicgao(usually, I’m out on the east coast this year) and I do have a node hosted in Chicago. I’m probably one of the few users with multiple seed nodes. With remote management of servers, it’s usually not much work that needs to be done locally, only when hardware fails, which if properly planned for, doesn’t need to be a big deal.
Ah, that makes sense - thank you 😀
I understand your choices to meet your needs, but those are not the same as Hive needs.
(Of course decentralized platform so it's just my point of view, but also kind of my field of expertise)
IMHO Hive doesn't need to pay for seed nodes. Witnesses are paid to run them.
"I encourage users to review their witness votes, and ask whether their witnesses provide public seed nodes."
And how do we know this? By asking the witnesses directly, or is there another way to check? [Sorry, I'm not technically savvy, hence the question].
Thanks a lot 🤝
By asking directly is a very good way (I mean, I guess it's recommended to vote for witnesses that you know that are real and actually answer your questions :-) )
Latest release (1.27.11) had those "initial" seed nodes compiled in:
seed.hive.blog:2001 # blocktrades seed.openhive.network:2001 # gtg hive-seed.roelandp.nl:2001 # roelandp hive-seed.arcange.eu:2001 # arcange anyx.io:2001 # anyx hived.splinterlands.com:2001 # splinterlands hive-api.3speak.tv:2001 # threespeak node.mahdiyari.info:2001 # mahdiyari hive-seed.lukestokes.info:2001 # lukestokes.mhth seed.deathwing.me:2001 # deathwing hive-seed.actifit.io:2001 # actifit seed.shmoogleosukami.co.uk:2001 # shmoogleosukami hiveseed.rishipanthee.com:2001 # rishi556Source: https://github.com/openhive-network/hive/blob/1.27.11/doc/seednodes.txt
Yes, I know. I know. But it's still difficult to get to know everyone.
So this suggestion helps. I already have a few from this list ;)
Thank you very much 🤝
Easy link to Rishi proposal - https://peakd.com/me/proposals/rishi556 :D
The proposal sounds good, I'm glad you're taking charge of the maintenaince of the servers. You have my support and my small vote for the proposal.
Best regards!
Every vote matters :) Thank you!
I know you mean well brother, and if I could I would try to help some more. I did however, bump you a bit on the witness list.
Thanks for all you do
Appreciate the support!
I've had the p2p endpoint enabled on my witness node running on a Mini PC. And the CPU usage has been meager even for that hardware. And I'm also running a VSC node and some apps on it as well.
I'm planning to make a public seed node with a domain, hopefully sometime soon. After the experience gained (and the continuing hived optimizations), I'm much more confident this hardware can handle it well. And I keep learning more docker features, it has some super cool abilities to limit the hardware usage of containers - this has been one of my main concerns in regards to potential DOS attacks.
It seems to me that if we learn and demonstrate how the network can run on meager hardware, this will contribute vastly more to the resilience. It will become accessible (both financially and in terms of required skills) to many more people to host nodes in their locations. And it will cost them very little, both as initial cost and ongoing electricity/maintenance (or necessary upgrades).
Got my support! GTG's feedback is decent, but I'm not opposed to this either way.
Congratulations @rishi556! You have completed the following achievement on the Hive blockchain And have been rewarded with New badge(s)
Your next target is to reach 6750 replies.
You can view your badges on your board and compare yourself to others in the Ranking
If you no longer want to receive notifications, reply to this comment with the word
STOPCongratulations @rishi556! You received a personal badge!
You can view your badges on your board and compare yourself to others in the Ranking
Check out our last posts:
Hi @deathwing / @rishi556, i sent 21 meowleo to meowcurator but it seem he only has 1 leo. How much time before i get my payout ? Been 16 days already..
It should have auto refunded you, did it not?
Oh wait I see the issue, you sent it to @meowcurator, it should go to @meowleo. I'll refund you now, and you can send to @meowleo.
Hello, thanks a lot.
But it seems like meowleo is empty too
@deathwing, can you refill it? Thx