@Criptix wen integrationnet?

Integrationnet is up and running since July :)

wen new spots integrationnet?

Most likely within the next months - the idea is that integrationnet validators get used to running nodes before they are moved to Mainnet :)

@criptix there's this dude called @Cryptotheos and I think he is ready for mainnet. He just posted a picture about his node which shows that something unusual is happening. It means that he can now face everything that comes the way of mainnet node runners

it would be my pleasure and honor

This is something the team will decide and announce, but my guess is that node uptime will be an important metric ;)

While I can understand why uptime would be a good metric for selecting the next batch of mainnet validators, it can also be difficult to get nodes back online to no fault of the operator. Some restarts happen lighting fast, other times they get stuck at waiting-for-download. Sometimes snapshots need to be cleared and redownloaded to pass waitingfordownload. I wonder how fair uptime is as metric for Mainnet selection given some of the issues well-intentioned node operators have encountered. For example, ive been sitting at my computer all day trying to get my node back up while others seem to have gotten online quickly.

I know it's sad to hear but the network just cares about uptime. That's what's needed in a node. If your node was the factor between life and death for the network, the network would die. That's why pro score is the way it is I think. Get that think tank on and see if any aspect of your set up could be improved.

Totally makes sense what you said and i agree - thus I mentioned it will be an important metric not the only one ;)

Yep at the end of the day the validator uptime is one of the most important thing for the network - because you need to be online to validate. Although for the integrationnet to mainnet transition I could imagine the activity of the node validator it getting the updates done and spending time on it will also go into the decision :)

I don't think I agree entirely. I understand that uptime will be the most important thing - eventually. I'm saying we havent been entirely in control of the uptime. When presented with WaitingForDownload, team leads have advised to "wait a few hours." After waiting a few hours with no change in state, we might be told to try to remove snapshots as a means to jump start things. While waiting, we can see other nodes getting back to ready/ready quite quickly. The point is, uptime (currently) isn't completely governed by the node operators right now.

Yep it is based on the nature of it being a testnet - so it's definitely a legitimate point 👍

Exactly! Anyway. Thanks for considering it. @criptix

Great point! Yea I have no idea what's happening beyond the node threshold. Wish I was there getting a look at how things are going. Sounds like some ironing out is in order. Do you think the specs or set up of a node could have any affect? Maybe a pro reset for those who switch to mainnet?

I'm won't pretend to be an expert, but I don't think it has much to do with node specs. Maybe if someone was running on janky home hardware/internet they'd be down more often. The issue I'm talking about is when the node needs to re-join the network after having been taken offline for restarts, updates, etc. It seems a little hit-or-miss how easily or quickly you'll be able to rejoin the network. Commonly, people have gotten stuck at a "waiting for download" state, which I think we're all realizing now that operator intervention is usually required to clear that state. With that said, it wasn't always so obvious. This should improve as the things are refined by the dev team. Also regarding node specs, I believe most of intergration nodes are run on D.O. and with nearly identical builds, so I doubt that's playing much of a factor today.

WaitingForDownload is an issue right now. It definitely has made it difficult to maintain our nodes. Testnet and integrationsNet is pretty lenient on uptime. Due to the fact that they are test environments and take more work to maintain at times. As long as your node checks in within every 72 hours it will be rewarded. But Mainnet rewards per snapshot. So uptime already rewards better/worse depending on uptime.

yeah. my original response to criptix was about if uptime should be considered for int > mainenet migration. I feel like it might not best / most accurate assessment of node op. performance today because of the issue you mentioned.

Probably not at this time I would assume.