I have some new nodes, which is on different public ipv4 subnets 1 to 1 NATed to the internal ipv4 address.
3 of the nodes is having good ingress but other 5 is having less ingress but also seems that they dont save as much data either? 63.7GB ingress and only 37.7GB user on disk?
How can i troubleshoot this? They are alone on each public subnet.
Ok did the success rate on a working node and a bad node here is the difference:
They are the same age ± 5hours
Hi @flwstern
Are all the nodes vetted? Ingress is noticeably different between vetted and non-vetted nodes.
I agree it’s odd how the figures across the board split the nodes 3 (doing well) and 5 (not so well).
I have another question related to your picture: Why does each node of yours have its own PRIVATE /24 subnet? This will not increase your ingress traffic. Only your PUBLIC IP will have influence of that.
Or is this purely for cosmetic reasons?
Nodes don’t receive an even amount of data. It is determined by a host of factors. Your node winning the race against other nodes. How close the uploader is to you aka latency to the uploader. Those are two main reasons, and there are minor variations as well.
Hi Stob,
Thanks for you reply, no they are not yet vetted.
For security reasons i dont want my nodes to be able to talk with eachother. If you read my post properly you would figured it out.
“which is on different public ipv4 subnets 1 to 1 NATed”
Hi Knowledge,
Thank you for your reply. They all have the same latency to the uploader/downloader. Cause they running virtual in the same VMware enviroment. SSD on OS and from same SAN storage. Same ISP.
Is there anyway i could troubleshoot the difference by the nodes?
Found the issue, my nodes that haves wierd traffic is having some sort of problem with their network nics or something. Getting bad download speed compared to the working ones.
Why are you over complicating all your nodes, They dont talk to each other so no reason to have them on there own subnets on the same network.
How do you map ipv4 adresses to different ipv4 subnets? Can you share this information?
Thanks and kind regards,
Hi,
So you own 8 x 17TB storage or you pointed all nodes to same volume?
I hope this is not storage in your work, that’s the best wish to get fired.
Besides, it will be many years before you fill up that 8 x 17TB.
Not if the nodes are all on different public subnets. At the current rate I’m seeing for a single subnet, those 8 nodes could fill up in just under 2 years. Of course over the course of that time I would assume data rates will just continue to increase so in reality they will probably fill up sooner.
My nodes are on my work network… but I promise I won’t fire myself. lol
it’s not possible to have the same latency to every customer in the world from different nodes and vice versa.
So, since nodes are different, they have different customers located in different places.
How can one determine if a node is vetted?
Not sure if there’s a way to check, but you will notice the ingress jump up once it’s vetted.
ah thanks for the link. turns out all my nodes are either 3 or 4 months old and only partially vetted.
in my case the vast majority of the traffic is coming form us1.storj.io, which is vetted, but has high downtime (like 2.5%). But some nodes just plain have higher traffic than others… i’m going to chalk it up to dumb luck. 
