An epochal epoch - Shelley update - 13 August

So, here we are. Nearly three epochs into Shelley as we approach the end of epoch 210. Incredibly, we have now reached the astonishing number of over 1,000 stake pools running Cardano. Sure, it’s just a number and just one measure of our journey so far. But what a number! It is a testament to the remarkable community that has grown around Cardano. The many many thousands of hours put in by our ITN and Shelley pioneers, the patience and passion of every ada holder. And our shared commitment toward building a decentralized future – together. It’s another truly significant moment in our story.

Talking of which, we’re just about to enjoy another…

This evening at 21:44:51 UTC, we’ll pass the epoch boundary and enter Shelley epoch 211. Shortly after the epoch has passed, we can expect to see the first-ever block created by a pool. The end of a fully centralized era and the start of our journey into the decentralized era. Good luck to every pool out there. Every block matters, but it will be nice to be the first. Of very very many!

d is for decentralization

This moment is happening due to a parameter change. Earlier this week, we submitted an update to the d parameter, that will hand off responsibility for 10% of block production to the community. By setting d = 0.9, the gradual process of block decentralization will have begun. This parameter will be decremented over time, thus continuing to decentralize block production across an already decentralized network of nodes.

We’re 5 days later than we planned. The decision to delay the decrement of d by one epoch was a very difficult one. And it was made very late in the day, close to the boundary – contradicting the guidance given in my last post. Apologies to all pools who felt – correctly – that we should have communicated this better in advance. However, very close to the wire, we felt it was only fair to hold off one epoch. We fixed the bug in Daedalus – which was leaving a sizable proportion of pools out of the tables. The extra time also helped bring a significant number of exchanges back online (see below). So while far from ideal, we felt waiting 5 days was the right thing to do. Thanks to all the community for your patience. You’ll find an updated delegation cycle visual below, reflecting these changes.

After holding d at 0.9 for epoch 211, assuming all goes well, we’ll likely lower d straight to 0.8 in epoch 212, thereby expanding community responsibility to 20% of all block production.

By dropping to 0.8 we can give more pools the opportunity to create blocks and establish themselves. At this level, pools won’t suffer in the rankings as long as they create one of the allocated blocks and get rewards. This way, we believe we can grow the network in a healthy way, at low network risk, while monitoring over the next few epochs before making any further reductions.

We’ll be publishing a comprehensive piece tomorrow on the IOHK blog, to outline the longer-term approach we intend to take in reducing d further. But in short, we need to strike a careful balance. We need to manage a steady, controlled and predictable rollout, while giving as many pool operators as possible (from the most established to the newest) the opportunity to mint blocks. On top of the original research, a significant amount of modelling has gone into this initial setting and we believe this approach is the best one, based on the data so far available. We’ll monitor the data over the next few epochs and if required, will make additional adjustments, including to k, the other factor important to stake pool ecosystem health. Keep an eye out for the post from Kevin Hammond tomorrow for more on this process and the plan.

Now you see it, node you don’t…

Earlier this week, we also released an update to the node, v 1.18.1. While it improved performance, today, we identified an edge case issue while doing some parallel testing on the Cardano testnet. So we’ve advised our stake pool operators to roll back to the previous node version 1.18.0. It will mean a bumpier ride over the epoch boundary which is disappointing, but we felt it was better to be cautious and stick with a battle-proved node for this important first block-minting epoch. Thanks to everyone in the stake pool community for being so responsive and rallying round on this. We’ll roll out a fix in a new version next week, along with a number of other scheduled improvements.

‘SMASH’ deployed

This epoch will also see the first deployment of SMASH, the service we have built after consultation with the stake pool community, to help with stake fetching stake pool off-chain data in Daedalus. Stake pool ticker symbols are not unique in Shelley. This has led to a few incidents of ticker ‘spoofing’ where bad actors have attempted to copy the tickers of well-known and established pools in an attempt to unfairly win some delegation.

While no ada delegator’s funds are ever at risk from duplicate pools, their ability to earn rewards on their stake may suffer if they delegate in error to a copycat pool and it performs poorly. This misrepresentation is not acceptable. SMASH stands for Stake pool Metadata Aggregation Server - and we have added the H at the end to make the acronym more fun :). Simply put, it is a database containing a curated list of stake pools, their ticker symbols and the rest of associated off-chain data. It will be managed and administered by the Cardano Foundation to help protect the interests of stake pools and ada holders alike, by referencing this validated list of pool tickers. It gives comfort to both and allows performant pools to build ‘brand’ equity with confidence in their ticker. Future iterations will allow additional community lists to be featured, thus broadening the options and decentralizing the service.

Exchanges continue to go green and Shelley-ready

Our exchange partners continue to make positive progress with their Shelley integrations. Inevitably, due to the various ways of doing legacy integrations, some partners are finding this process more involved than others. Our dedicated joint IOHK/CF exchange relations team continues to work hard to support all our partners through the process. You’ll always find the most up to date status information we have published here. You may also find further updates on your preferred exchange’s site. Currently, around 40% of ada has been staked and we expect this to increase over the week ahead as more exchanges get Shelley-ready. If you continue to have issues depositing or withdrawing ada from your preferred exchange, remember you can always open a support ticket with us. While the responsibility for integration ultimately sits with the exchange, our team is always on hand to lend a hand if required.

That’s all for now. Good luck to every pool operator out there – may the first (block) be with you! Thanks as ever for the support and we’ll see you back here again soon.

22 Likes

was able to roll back 1.18.1 just in time to see the epoch change lol.

thanks @IOHK_Tim, but how were stake pool operators “advised” to roll back to 1.18.0? All we saw was 1.18.1 marked as the “latest release” on Github (it’s no longer appearing as a tag or a release) and therefore we upgraded ASAP. I appreciate the notification here, but this is a marketing message which we might not have read. Can you please indicate the standard place where future version recalls & upgrade notices will be posted?

2 Likes

Not involved myself but I think you have to follow “live” channels rather than just rely on github. This information was on Telegram, Twitter and other channels, admittedly it was very short notice and they’ve apologized for that. If I was an SPO I’d be on Tw and Tg as an absolute minimum.

2 Likes