Little glitch in the dashboard makes 40K nodes appear again.
That’s one whole Bitcoin blockchain that got rediscovered 
As mentioned in the other post. I am really sitting on pins and needles right now waiting for the second post about how Chia lost those nodes 
Nice! Still one of the most decentralized blockchain.
So, maybe there was no “little glitch” there afterall.
Chia nodes chart in the lower left corner - Grafana
Looks like Chia doesn’t like to do post mortem. Either it is embarrassing for them, or have no understanding what happened. Sure, “why bother” is the third option.
So, just a couple of hours ago, Chia lost around 60k nodes (hard adjustment). Maybe that indicates that what was perceived to be found before was just a happy error. Either way, makes those charts to be kind of unreliable (contrary to how they were advertised).
If those charts are going to be some kind of “the primary goal of these dashboards is to provide an accurate and third-party verifiable accounting of the blockchain and its state,” maybe it would be worthwhile to pay extra $50 per few amazon instances, and make them as robust / redundant as possible.
Also, we can assume that there is a strong correlation between the net space and the number of nodes out there. Sure, some of those nodes represent nothing (secondary nodes, or nodes run where only harvesters would suffice), where other represent whales. Still, seeing ~40k nodes found before, and ~60k lost this time, where net space doesn’t flinch, rather suggests that there were no nodes either found or lost.
2 Likes
Yeah I askled about this in Keybase and got this reply:
they didn’t fall off… I think its likely because new dns introducers just replaced the old ones, and now have a slightly outdated DB… Should catch back up
should have had a newer DB to start with though, so I’ll figure out why it didn’t
But yeah, it doesn’t help the reliability of this data. I’ve also wondered about the node count staying level while netspace is steadily dropping.
If you look at below, that weird bump in nodes does seem weirder than before
1 Like
Also as a note, you can get the peer count yourself with the crawler i’ve been told. haven’t tried it yet but I will.
I think that should be this:
As that is a crawler, it may miss those nodes that don’t have port 8444 open.