@0xMert_
@TrustlessState
with inflation going down and number of transactions going up- take the long term inflation rate of 2%, solana will actually be deflationary at 50k tps. I don't blame him for thinking 50k tps is not needed or possible because he hasn't seen and can't conceive of more than 20 tps
Today we jumped our Solana validator overseas to Tokyo and are now running the
@jito_labs
validator client. Here's to geographic decentralization, MEV rewards, and a fast new server with
@latitudesh
🥂
@0xMert_
@TrustlessState
solana will be deflationary when it deserves to be deflationary and we have apps that are making use of abundant blockspace. unlike eth which wants to artificially limit it to extract more fees and prevent apps from being built, because they care more about short term circle jerk
Greetings fellow Solano comrades who have made it to Solana Summer '23.
About one year ago, we (
@7LayerMagik
and
@dubbel06
) started Overclock validator on Solana and decided to go with the oh so loved, but oh so hated 0% commission.
@TrustlessState
@R89Capital
this is completely wrong. solana has the same features eth does, but they're feature gated because
1) demand for the feature (eth NEEDS L2s to function, so doesn't)
2) currently, it would hurt performance
everything eth gets from "zk" is because of 3 precompiled alt_bn128 calls
@aeyakovenko
@thal0x
@solana
I feel like there should be a highly publicized, incentivized battle test for this. If it doesn't pass... it gets pushed off again for a number of months until it can finally get the black belt.
Today we had ~80 minutes of downtime. Unfortunately our primary server and 1st choice backup server stopped progressing at around the same time :/
The issue seems to have originated from spam and was remedied with more conservative firewall rules.
This morning we migrated our validator to its new home in the Zag1
@digitalrealty
datacenter in Zagreb, Croatia.
We wanted to help further geographical distribution and it is currently the only validator located in the beautiful country of Croatia.
We have implemented our Protected Staking Rewards (PSR) bond for
@MarinadeFinance
and are looking forward to seeing this innovate idea play out!
PSR will protect stakers' yield from validator downtime and goes live in a few hours in Epoch 608
📣 Attention Solana Validators, here are two big updates to Protected Staking Rewards and the new delegation strategy:
1. The new Marinade delegation strategy takes effect in epoch 601. You can visit the validator dashboard to see your new planned stake.
We gave 1.16.9 a shot for the past 2 days and the RAM decrease is quite amazing (see below) but there seems to be an issue for some validators where servers are crashing at times. After experiencing 2 of these ourselves we have downgraded to 1.14 for the time being.
Overclock Updates - March 2024
1. Next epoch (596) we will be updating our validator to achieve higher vote credit performance through non-lagging voting methods.
You can find us on and if you have any questions about this let us know in Discord~
We believe Diet Clients (a form of trust minimized light client) are the next step in decentralization for Solana — and that means we'll need a system where users can verify the chain both cheaply and easily when needed.
Despite testing things ahead of time with a test validator, we ran into some unforeseen bandwidth issues in Tokyo which would have led to significant cost increases. We have thus decided to move back to our home state of Texas for the time being until we can debug further🤠
Today we jumped our Solana validator overseas to Tokyo and are now running the
@jito_labs
validator client. Here's to geographic decentralization, MEV rewards, and a fast new server with
@latitudesh
🥂
Despite testing things ahead of time with a test validator, we ran into some unforeseen bandwidth issues in Tokyo which would have led to significant cost increases. We have thus decided to move back to our home state of Texas for the time being until we can debug further🤠
We'll be updating to the Jito client in the next week or two, once Jito's New York block engine is up and running. Brainstorming regarding MEV revenue sharing for our validator will commence soon. Many thanks for the opportunity to participate in the 1st stage of the stakepool🙏
We just launched JitoSOL, Solana’s 1st MEV-powered liquid staking derivative!
The token offers best-in-class yield while improving Solana's decentralization and network security.
Users can swap SOL for JitoSOL now at
Is the voting fee reduction still on the table
@aeyakovenko
? Sorry for the Sunday ping, just currently trying to figure out whether we need to reduce server costs soon or hold out until a potential voting cost reduction. I guess a base fee/cu decision needs to happen first?
Just bought ourselves a 2nd server (Epyc 7502P) so that we can eliminate downtime, including for validator update restarts that hurt APY. Thanks to mvines for engineering the seamless failover process and
@backbone_link
validator for working to make this process even easier :)
Running at 0% commission helped us gain enough stake to finally overcome our cost burn earlier this year.
We greatly appreciate the stakers who have supported us. The past year has been rough, but the Solana community is tough and we plan to be here for the long haul with you.
Our application has been submitted~
We would love to take part in this and inject some of our collective brainpower into what will be a critical on-chain governance entity for the Solana eco
A little bit of alpha on our own plans is also included :)
1. The Jito Foundation is pleased to announce a public Call for Delegates!
Individuals and organizations that wish to guide the future of the Jito Network should follow the application instructions below.
@vega_gx
@SolBeachBum
@antiprosynth
life is fun when you're ignorant of how liquidity works and forget that you need all that TVL to prevent slippage. phoenix does higher volume with less capital locked because they're not a constant product AMM- which is basically all there is on those l2 chains
@TrustlessState
@R89Capital
additionally, solana HAS considered other cryptographic primitives as well, such as KZG commitments in turbine, but guess what -> the constant proof size is not worth the performance against just using merkle trees
What Shaun will be contributing to is important for Solana and something that we will describe in more detail in the coming weeks.
Special thanks to
@fd_ripatel
for helping us to make the connection with Shaun (shaunuk on discord).
@0xdoug
> without the overhead of execution/block building
solana is already working on bankless leaders and APeX (async prog exec).
and yeah, you're right. reason people don't do it is because they're still stuck in the dumbass narrative that solana is only fast due to hardware
We went with "No" for this SIMD although we are willing to vote for it in the future. We would first like to see:
1) A native block reward distribution mechanism so that this process is more standardized. Reliance on non-standardized ways to pass along these rewards might
Two epochs left in the governance vote on SIMD-0096 - Reward full priority fee to validators
Remember to cast your vote and be heard, all validators can vote!
Proposal
Voting instructions
Current status: 21.16% of stake has
We have had a similar issue once again. Our server is back up but we are still in the process of diagnosing the issue and figuring out future mitigation steps.
We will be lowering our inflation commission to 0% for the current epoch (614) along with the next one.
Today we had ~80 minutes of downtime. Unfortunately our primary server and 1st choice backup server stopped progressing at around the same time :/
The issue seems to have originated from spam and was remedied with more conservative firewall rules.
In order to professionalize our business more, add extra infra if needed, etc, we will be increasing our commission to 4% in approximately 1 month.
We also intend to upgrade our contributions to Solana and make way for smaller validators to gain stake with lower commission.
Lastly, if you have any questions or concerns, please reach us in our discord
You can also monitor our commission and performance with
@laine_sa_
's alerts:
Thanks again Solano friends and onward we march🫡
@SolBeachBum
the beauty of this tweet becoming popular is that
1. engineers, builders obviously know its troll and will continue building on sol
2. idiots who don't do research but just buy tokens will fade it.
widens asymmetry, improves the upside for someone who does basic research
So how do we plan on upgrading?
Something on our minds for a while has been the desire to have an even stronger band of grassroots contributors to Solana core R&D, perhaps best unleashed through the latent power available w/ staking and validator culture.
In alignment with this aim, we will be adding a new team member, Shaun Colley, who isn't active on CT, but who is a badass community engineer that has earned various Solana bug bounties along with contributing to Firedancer.
For reference: . "This document demonstrates how to transition a staked validator from one machine to another without a restart. Using this technique, a validator software update can be accomplished with virtually no downtime."
@aeyakovenko
@thal0x
@solana
Zan wrote something like this for testnet. The Solana "hammer" which creates a bunch of large "hammer" accounts. I just think it would be good to make a public spectacle out of it. Maybe even
@TheBlock__
could run their own spam node
We migrated our validator from Tokyo to Prague earlier today and the DC has already had an issue that caused us downtime in the last hour. If the issues continue we will migrate back to our Tokyo server.
We plan to move to a new, more permanent location in the next few weeks.
We've set up two
@heliuslabs
Atlas transaction senders/forwarders in Tokyo and Zagreb. Hoping the stakeweighted QUIC forwarding we're adding here improves the chances people land transactions.
@TrustlessState
@R89Capital
you can check all the feature gates here: and there really was demand for it, it could be activated - just need to hop on the core dev call and raise your hand, make your case
@SolBeachBum
i really don't understand why eth people talk about zk in relation to scaling at all. it does nothing to help you scale throughput / latency
zk tech would help your prover scale.
provers work on an execution trace.
an execution trace is produced AFTER your node executes the block
Announcement! Our friend
@ejwallach
will be joining the Overclock team to head the
@SuiNetwork
arm of Overclock with
@dubbel06
also helping with infra operations.
Eric has a background in fintech/web3 investment and is a good follow for those who love Dune dashboards.
Overclock Sui is proud to secure
@SuiNetwork
! Our primary team members
@dubbel06
and
@ejwallach
contributed to Sui Testnet Waves 2 and 3, as well as the Sui Mainnet genesis ceremony, and have participated in the Discord since April 2022.
@buffalu__
might be an overlap, so apologies in advance but
- worth mention how CPI is different
- no SSTORE, so need to get familiar with the concept of account rent
- no concept of querying. in evm you can call a contract to get some of its data, but its fundamentally separated in sol
The issues are currently under investigation by Labs and other ecosystem engineers/validators. Looking forward to moving back to 1.16 when the underlying issue(s) are identified and fixed up. Only a minor % of mb is currently on 1.16.
MARINADE'S DIRECTED STAKE IS OFFICIALLY HERE!
You can now stake SOL with a single validator and get $mSOL for Solana DeFi. And it's officially LIVE!
🧵How it works (1/10):
@leDuckTape
@0xMert_
@TrustlessState
forking the evm and forking the svm have the same effect - my personal thesis is that sol's decentralization and scale does matter - it can sell blockspace to whoever needs it, it can be apps or it can also be L2s (sol can be a better DA layer than eth or any others)
Hey everyone, Overclock's commission has been raised from 0% and the epoch will conclude shortly.
If you have stakewiz's commission alerts set you will have noticed that the commission is actually 3% now rather than 4%.
@arixoneth
> You do realize that almost every blockchain scales with hardware?
no it doesn't. scaling requires engineering effort. writing scalable code precisely MEANS writing code that scales with hardware and exploits it to its fullest. EVM doesn't even exploit multiple cores
@noah_pravecek
@aeyakovenko
you don't need DAS if you don't need light clients to get an assurance that data hasn't been with-held. if you're running a solana full node, you get assurance that data has been published because you download the entire block.
@antiprosynth
@vega_gx
@SolBeachBum
which is the point. you're perpetrating a system where the eth-usdc pool keeps giving away money for free to arbitrageurs. all you do is show these metrics to assert some kind of dominance when it's clearly not. the eth-usdc LPs LOSE money. but hey, if it pumps eth who cares?
@vega_gx
@antiprosynth
@SolBeachBum
he fundamentally doesn't understand the purpose of liquidity or how to measure it. for him, it's just some "metric" he can push to show dominance wether people actually trade there or not.
there are better measures of liquidity on a clob such as a round trip loss
@toghrulmaharram
@metaproph3t
@TrustlessState
i think the bigger solution is actually state rent which allows the state market to grow separately from compute market.
in evm, the use of SSTORE op to price storing inadvertently leads to current users having their storage cost subsidized by future users
@leDuckTape
@0xMert_
@TrustlessState
forking the svm can give you fast execution but it does nothing for scalable fork choice, fast data dissemination or even scalable data dissemination (turbine doesn't care how many nodes are in the network). Ultimately, sol's blockspace has value in being the most scalable
@aeyakovenko
Zan did a calculation here btw and with perfectly even stake (very unlikely we'll get close) that would be 39.4k stake per validator with 10k validators. With 10% commission each validator would be losing $7.5k per year. I'm not sure what the assumption here is on server costs.
How will you deal with the issue of duplicate txns sent to separate block leaders
@aeyakovenko
? Either they have to constantly pass info back and forth as block is built (more latency?) or they need to be bankless... and won't know if other leader has processed same txn?
2. In our future upgrades we will also be implementing "seamless upgrades" where validators can upgrade and restart their servers without downtime.
Michael Vines,
@pumpkinspool
, and
@ShinobiSystems
have contributed a lot of know-how to the ecosystem regarding this process.
@SolBeachBum
l2s are often overrated (all except 2 of them are missing features to even be called real l2s)
sol is underrated (clowns legit believing there is an "on-off" button or that ftx had any influence at all on the technical decisions or the validator community)
@0xcarnation
@Crypt0Panda
@ledger_top
50% priority, 50% base are both burned. also the dash you show says "non-vote" but even vote txn fees are burnt 50%. you can calculate the amount of tps needed to be deflationary. 13k tps causes roughly 1m sol per year burn (excluding priority, state). this could change though
@sobanraza_
@JosipVolarevic2
its a permissionless system ser. you can build whatever you want to. plenty of useful products trying to find PMF on sol.
also, not all forms of speculation are "scams"
@cryptomiester
@solana
alt_bn128 syscall will be glorious.
@0xMert_
@7LayerMagik
if true, can verify certain snarks, kzg commitments, threshold signatures. eth gets a lot of mileage out of that precompile
@0xMert_
@AlexMasmej
what's funny is that the proof system has nothing to do with how much throughput the rollup can have. proving is after the fact once you've already finished execution. risc0 or valida getting faster doesn't mean your single threaded, bottlenecked vm can suddenly execute faster
@dubbel06
has enabled our MB Grafana metrics. Replay times are looking fast on an Epyc 74F3... Execution times appear faster than all Lido nodes that I've thus thus far compared against, along with some other stats.
@noah_pravecek
@aeyakovenko
in terms of block dissemination, solana IS the fastest, so it makes sense to use it for DA. Now how you go about that is a different issue -
1. memo program (like call data in eth)
2. account (horribly expensive)
3. compressed state (my personal favorite)
@metaproph3t
sorry, but almost done with the rant - if sol is not being used for DA or not supporting that use case, then sol is severely under utilizing the value of infra that it provides (not exploiting turbine, PoH etc). nothing wrong with that...just not as efficient
@diyan_slavov
@toghrulmaharram
@StrategicHash
@SolanaLegend
@jon_charb
it'll be challenging and perhaps there would need to be incentives for client diversity.
overall, I would imagine there would be a significant performance benefit (not as fast as every node running firedancer. at least not until a time as labs client in rust is also re-written)
@metaproph3t
idk what the cases are for saying eth is better at those than sol (other than some eth is settlement bankless dogma). it's not. sol is cheaper, has a higher capacity for DA and offers quicker finality for l2, so based sequencing is more efficient
@diyan_slavov
@toghrulmaharram
@StrategicHash
@SolanaLegend
@jon_charb
not at the exact same throughput, but several of the decisions being made by firedancer are also possible with rust. eg: numa + pinning process per core with a spinlock... this can be done in rust as well. algo changes for ed25519, rs-gf8. firedancer has the benefit of hindsight
@disk0x
@0xMert_
its not about transaction ordering. its about block ordering.
in a fully synchronous protocol
leader n produces block > validators vote on block > leader n+1 produces block off of that
with PoH
block production is disconnected from consensus and execution.
Just a reminder that in roughly 4 days Solana will reach Epoch 481, which is when 4% commission will turn on for Overclock. You can track epoch progress at the top of .
Let us know if you have any questions!
We're happy to collab on this and thanks to
@earthyfrodo
for already taking a stab at some improvements. Let us know in our discord if you have any issues or suggestions.
4. For those interested in staking with Overclock but want the ease of dealing with a LST instead, we recently launched the single-validator LST, clockSol, through
@sanctumso
.
*Note that LSTs differ from native staking and have their own pros/cons
1/
@OverclockSol
's clockSOL
The Overclock Validator team have massive Solana expertise, constantly involved in high-level discussions to improve the network.
clockSOL holders get the benefits of having some of the biggest gigabrains in the space, your staking returns is
5. The last two months we have been sharing quic connections with
@heliuslabs
in the hope of improving the UX for Solana users during these wild on-chain times 🤠
This week we are adding two more Atlas forwarding servers (EU and US).
@metaproph3t
no, thats not necessary - you need proof of data availability to ensure that data has not been with-held. data custody is an orthogonal f+1 problem that's solved by IPFS, arweave or any DHT
@metaproph3t
one good example of this that we're looking into is potentially using cnfts or compressed state for this because
1. commitments are stored permanently in accounts and efficiently
2. offchain indexing protocols already exist for this. the offchain data here is just blocks
@STEVEW00DY
@Siggy_SA
@staratlas
@solana
bot accounts is not a bad thing - they're paying transactions fees. it's the same for oracles, market makers and arbitrageurs. blockspace is there for whoever wants to pay for it.
also star atlas took botting into consideration during design (they reward vertical progression)
@metaproph3t
yes, duplicates are an issue, but you can build on top of that by adding bits and pieces of a protocol such as det hashing to determine which txn goes to which sequencer, even preferred sequencers so that you can execute faster since you know certain blocks will be ordered first
@trident_of_eth
@tofushit888
@circle_crypto
what active defi ecossytem? xyk swaps that lose money without sufficient tvl and people arbing them because that's necessary for price discovery isn't "activity" any more than all of polygon's wash trading on NFTs
3. We received a grant earlier this month to work on what we believe will be an important infrastructure piece for the Solana ecosystem. Our goal is to reduce the hardware requirements needed to fully verify Solana.
We'll share more about this when development is further ahead!