Is this true? @Alexey
Yes, ideally, but that doesn’t happen. Right now, this (non)garbage is simply not taken into account when calculating storage on the satellite side.
I doubt this. I don’t think the restored from trash data has some markup in the satellite db. If it’s restored from trash and is in the data blobs, were the sats knows it is, than it’s paid data.
Storjlings could bring some insight…
It’s partially paid. The main reason for restore usually when the BF was wrong and moved not deleted pieces to the trash. So, these pieces will be paid, but the deleted pieces are not paid.
We do not have an efficient mechanism to restore only pieces which should not be deleted, we simple do not know, and time is limited to figure out each piece. The restore function doesn’t support restore only some pieces.
However, we also have an integrated mechanism for downloading, auditing and repairing, this node can recover a piece from the trash if it was requested and notify the satellite that this piece was recovered from the trash to alert engineers and figure out what the problem is.
1 Like
So worst case one of the estimated 100 million files you restored from my nodes is paid. 
1 Like
Unlikely, this function is called when the number of restored pieces from the trash during any GET* method is great enough to conclude that the one of BFs were broken.
So, more like hundreds.
I think storj have to pay for their mistakes. Is the space allocated and in use? Yes. Do we get paid for it? No. Is this charity?
2 Likes
I’m not obligated to decide. I described how it works now.
Of course, I have shared your opinion, thanks.
1 Like
The question is how big is the earnings impact? If the node is not full, there should be no impact, cause you’ll receive still data. But if it’s full, then it’s like half a month lost on ingress.
But ultimately I think restored trash should be treated as paid storage until the restore function was triggered. I think that would be fair. But the decision is in storjs hands
But I don’t know how hard it would be to implement such a Funktion.
I doubt that the restored trash is 1TB per node, but even so, you are not paid 1.5$. Wow!
And that’s if it stais the entire month.
Just let them sort this out and rectify the bugs.
The most important thing is not to loose clients pieces. Stop bitching about every cent you are not paid.
Nobody forces nobody to find bugs, to report bugs or to run nodes.
1 Like
Almost no one paid attention to such problems while the income part was higher. It’s business, nothing personal. Now, in my opinion, there are many more problems, and payments are steadily decreasing.
3 Likes
Not arguing with this part, but…
Are they? I do see slight growth after the great purge and the subsequent performance tests.
1 Like
Maybe in absolute values. I always calculated through the effective cost of storage per 1TB. Right now for me it’s ~$1.47 total (storage+egress) per 1TB of used space or ~$1.35 (only storage)
Then… I don’t observe it either. I’m consistently getting 1.3–1.4 USD/TB, counting uncollected and collected garbage in, for a long time now… maybe with temporary peaks when there was less activity in the network in general, leading to less lingering trash.
2 Likes
Theyre prolly referring to the “good ol days” when we were paid unsustainable amounts before they dropped the payment rates. My nodes were in the $4-$10/TB range. It made no sense. But they paid a ton for egress.
1 Like
You are missing out nothing. If you still have space the ingress continues as normal. And even if the disc is full, what are you missing? For me it would be a whooping 1 USD for 1 Month per 30 USD pay (and my nodes are not full). So nothing really lost here. And if you are really struggling because you are missing out on 1 USD then you have other problems.
I like that some people creep out of their hole just to shit on this project, because there are some (not even major) problems, that needs to be checked.
3 Likes
If we accidentally start an “unpaid trash” discussion every 4 months or so… and accidentally get some hole-people upset enough that they actually leave… that would be an unfortunate accident.
(sets reminder)
6 Likes
For the 5670 full nodes it is a loss. For all other nodes it actually means more ingress.
I have no full nodes atm… 
Not necessary. I have 2 full nodes which are backed by 2 other (not full) nodes. And I think many that have full nodes have a secondary node running or are even with collected trash at their limit. (My full node has a capacity of combined 3 TB and have less than 100gb Trash)
Edit: where can you see how many nodes are full?
There may be an easier place: but you can search the raw stats for “full_nodes” and just use the highest number (since most SNOs are on all four satellites anyways)
3 Likes