The difference is that your nodes could have pieces of the same file on one PC instead of five.
So, with your PC failure the network could lost 5 pieces of each file in the same time.
But @BrightSilence is right, we should add to restrictions your case too
thats why i think you should allow make more nodes on single ip without bandwidth restrictions, but restrict satelite to write more than 1 piece to this nodes.
2 Likes
Today people make striped partitions to make big nodes, i made nodes with 1 hdd, per node, hdd dying is the bigest posibilitie, other can fix
Itās not a bandwidth restriction. It is exactly restriction to store pieces of the same segment on /24 subnet. Those pieces should not be in the same PC or even group of PC in this area or with this ISP.
In short term it will looks like a bandwidth restriction, but long term your nodes can store more data than a one node and could have a more egress because of that.
Your circumstance is just a dangerous setup for the network.
1 Like
But you exactly restrict Bandwidth, as i mede separetly i geting exactly more information.
1 Like
Thay have same ISP, this mean i shouldnt have same pieces.
If your IPs in the same /24 subnet
This solution is not ideal, but it can make it difficult to concentrate a huge amount of data.
I wil think about deviding them for some parts and locations, but on other hand i can mantain them here much faster than in other place. And here i have server platform pc, UPS.
1 Like
Thank you for understanding!
Iām almost 100% sure heās not the only one doing it Iām sure a lot of others are also doing the same to max profits if storj wants it to stop they have better put something in place to pervent it.
1 Like
If one node goes offline, the customer will not even notice. Keep in mind that each uploaded chunk is split up, 80 pieces are stored across the network, but only 29 are required to reassemble. If a single node goes offline, 1 out of 80 pieces are not accessible, but the customer only needs 29 so this is not a big deal ā 79 are still available.
This is why the network only stores 1 out of 80 pieces within the same /24 subnet, to geographically distribute the pieces and lower the odds of multiple pieces disappearing at the same time. (Losing 52 pieces of a chunk means the chunk itself is lost.) If you are circumventing this mechanism using proxies and your system totally fails, then up to 5 pieces could vanish at the same time. The more nodes you run this way, the more of a problem it is for the network as catastrophic failure of your ISP/server or even a natural disaster could wipe out dozens of pieces of the same file.
If this happens too much, people will stop trusting the network and using it to store data. And if that happens, your nodes stop making money.
3 Likes
Hey @Vadim,
I could be wrong, but it seems you have flipped the upload and download terminology in this tool. Storj has throughout their network and logging consistently used upload to mean an upload TO the network and thus an upload TO your node. Download is a download FROM your node. This is how log lines are displayed for node logs as well.
On the dashboard it does display traffic from the perspective of the storagenode, but intentionally uses the terms ingress and egress to prevent confusion with the terminology in the logs.
It seems your tool translates ingress to download and egress to upload. Which is reasonable enough in isolation, but really confusing in the context of Storj. I recommend sticking with ingress and egress as terminology, or flipping your translation to align with the terminology used network wide.
Thanks for feedback, will fix it on new version.
1 Like
I fixed Egres and Ingress, also aded monitoring ability to discord @BrightSilence can you try it also?
On my end looks like working. I send you personal message with deteils but whatch last relese.
i tell this before v3 started. Entering this restriction simply hides the actual location from the network in other words, the exact opposite of what the creators expected.
Looks good now! Thanks.
I donāt really use Discord and have no need to have an additional way to monitor these things remotely. So didnāt test that part.
OK, Ichosed discord, because it has free API and, have app to phone, i can almost online to know if my nodes offline, will get message to phone.
Just shrink it to right all will be.
No, i canāt expand the window any further to the right. The right column seems to be fixed width and ALWAYS looks like in my screenshotā¦