Hey, so I have a weird situation. It’s not really a problem, but it is really weird behaviour.
I have a node on my PC that is 9 months old (fully vetted) Im gonna call it #1 node. It’s on an external drive.
I also added 2 new nodes on my PC. They are new nodes, a few days old. They are on a different external disk.
Here is the ingress they got today: #1 = 20gb -100% ingress #2 = 10gb -50% ingress #3 = 10gb -50% ingress
I just want to point out, that #2 and #3 aren’t vetted at all, so they should get less ingress. Also for some reason #1 node is not splitting its ingress with other 2 nodes.
Also when I was adding those nodes, when I had only #1 and #2 running, both were getting 100% ingress. The same amount. But #2 wasn’t vetted.
#1 ingress traffic didn’t change after I added #2 and/or #3. But #2 ingress halved when I added #3
These 3 nodes are on the same PC (IP)
As far as I know they should split 33/33/33 of ingress, right? But nodes that aren’t vetted should get even less ingress? so maybe it should be like 50/25/25?
It’s not this simple and is actually very hard to calculate as it depends on a lot of external variables. It’s been discussed before but comparing your vetted node to unvetted isn’t possible. This thread explains it further:
Firstly, you need to consider each satellite separately. Then you have a set of vetted nodes for a given satellite, and a set of unvetted nodes. All vetted nodes for a given satellite split vetted ingress. All unvetted nodes for a given satellite split unvetted ingress. A node might be vetted for one satellite, yet still unvetted for another.
We recently observe that per /24 block the unvetted ingress is around 30-40% of the vetted ingress. Let assume that your old node is vetted on all satellites, and your two new nodes are unvetted on all satellites—this is pretty likely given their age. This means that your two new nodes should be roughly equal in ingress (which is what you observe), and they together should be less than half of your old nodes, so the proportions should be around 6:1:1.
Also, if you look at traffic as reported by the node’s web console, the data for the last day on the diagram may not be up to date. It’s best to compare traffic for one, or maybe even two days before the last one.
If I understand you correctly, what you observe is more like 2:1:1? This may mean that there are other vetted nodes in your /24 block, maybe managed by another customer of your ISP.
Yeah, if you had read my comment, you’d notice I placed an assumption:
What you suggest is that this number got to 100%, i.e. there’s 20× more vetted nodes with free space than unvetted. I do not observe that on my nodes yet, though. I do admit the last time I had a fully unvetted node was a month ago, but I would be surprised the situation changed so quickly.
I know I reported this observation here, but I’m somewhat certain¹ that other people here also reported it on the forum. I can also say that a month ago this observation was still true.
Please note I wrote per /24 block, that is, there’s far more /24 blocks with vetted nodes than those with unvetted nodes.