this is needed only if you have also errors like “database is locked” in your logs. Otherwise it’s not required, but may save some IOPS though.
Thank you! I’ll see if the above changes rectify things and then consider this later if needed!
The defrag and removing index are taking a long time!! Was one of the original replies correct when they said that it can be offline for 10 days without repercussions?
Also, on a slightly separate note, I’m keen to hear your thoughts on profitability:
When I first joined, I don’t think there was any easy way to ‘scale up’ eg I don’t think multiple nodes were allowed from the same IP and I think adding additional drives was also problematic.
Is this still the case?
In short, is there any way to make substantial money from the project? Or should I continue to look at it more a hobby style project?
Yes, the node could be offline for the 12 days, then it will be suspended. After that you would have a week to fix an issue. If the node would be offline or suspended after 30 days, it will be disqualified.
We implemented an /24 subnet rule, this is mean that all nodes behind the same /24 subnet of public IPv4 would be treated as a one big node for uploads (i.e. they would share the same traffic, as if there would be only a one node) and as a separate ones for downloads, repairs and audits. This effectively spread the risk of losing a one node if you have multiple, in that case you will lose only part of the data (the one node from several), not the whole data, as if there would be only one node.
It is also mean, that running more nodes will not give you more income than the node of the same size, so no point to run many nodes on the same disk/pool. It’s useful only if you have a spare HDDs with a free space and do not want to learn or configure RAID with a redundancy.
See also RAID vs No RAID choice
1 Like
Thank you for your reply. Just to double check I’m understanding correctly, would you mind confirming my understanding:
The purpose of setting up additional nodes under the same IP is to act as a backup - both nodes will store the same data, but be audited independently. Essentially, it’s an easy alternative to setting up a RAID structure. Is that correct?
As for monetisation, does this mean that, given my current setup of one 18TB HDD, there is no additional monetisation I can do, even if I were to add additional 18TB HDDs under the same IP? Or am I misunderstanding that part?
Or, to put the question in another way: as my 18TB HDD is full, what next steps are available to me to increase my profitability?
Thanks so much again for your help and patience with my questions!
No, they would store absolutely different data. They would just split the incoming data almost equally (after we implemented choice of 2, one of the node could have more ingress if it’s more performant than the second). So they are not a backup. However, if one node die, you will lost only that part of data, but remaining will still work. If the remaining node is only one in the /24 subnet of public IPs, it will receive a full amount of every new data, not the half like when there were two nodes (or 1/3 if there are were 3 nodes, etc.).
All incoming traffic will be split between these two nodes, while they have a free space. As soon as 1TB would be filled, your second node would get the whole incoming traffic, not the half as was before the 1TB node become full.
The only thing, that should be taken into consideration is to run the next node only when the previous one almost full or at least vetted.
Because each new node must be vetted on the satellites, It can receive only 1%-3% of the customers’ uploads until got vetted. To be vetted on one satellite the node should pass 100 audits from it. Usually it should take at least a month for the one node in the /24 subnet of public IPs. Since audits frequency depends on amount of stored data, more than a one node could be in vetting process much longer than the only one.
1 Like
That’s great, thank you again for your patience in explaining this!
So, to confirm:
- We can have multiple nodes
- Incoming traffic is split among nodes
- Recommended to allow each node to become full before adding a new one
- Each node is score independently, so if one fails, it doesn’t effect the others
This is great!
So, final questions from me:
- What is the maximum size we could grow to?
- Any thoughts, recommendations or considerations I should be aware of around trying to do this to generate income? e.g. what is a wise/realistic income goal to aim for?
- What are the consequences of a drive failing and losing the data?
Thanks again!
yes you are correct.
there are none. But the usage depends on customers, so the more amount your node have used, the more deletions would happen. So at some point they would become almost equal. I do not know what’s the current equilibrium point.
Our documentation suggesting to provide 24TB per node as a maximum recommended size, however it’s not a requirement, it’s a suggestion.
We do not provide such advises. We always recommend to use what you have now and what will be online anyway, with Storj or without. In that case all income is a pure profit.
You may use this Community Earnings Estimator to get an idea how long it may take to fill all the provided free space and how much you may earn:
The node will be disqualified by each satellite independently and the held amount on the satellite will be used to recover missing data on other nodes:
If your node would be disqualified not on all satellites (you were able to recover almost all data), you may continue to run the node, all other satellites will still pay for the usage. If the node is disqualified on all satellites, you need to start over: delete the current identity, generate a new one, sign it with a new authorization token and start with a clean storage.
Thank you so much! This is all really very helpful, I think I have everything I need to proceed with adding another harddrive 
One final quick question, if at all possible -
Is there any ‘hall of fame’ style information anywhere showing the largest nodes currently operating?
The basis of the question is just trying to get a real understanding of what the upper limit of possibility is with my overall node (how big can a node realistically grow).
Thanks!
1 Like
no. but why do you need it? You should be the best, that’s all.
“the best” is qualified as “filled 100PB after a day”. Otherwise, well 1 of 20k unfortunately. So pointless in my opinion… but well…
There is no upper limit. Unless your node would meet an equilibrium point (when the amount of uploads almost equal the amount of deletions), it’s unique for every node though.
So just try it.
Thanks! The reason for my question was to establish how realistic it is to achieve certain milestones. For example, if no one has every made it past 1PB but many have made it to 100TB, then I can get a better idea of how realistic this is and, therefore, how much brainspace and time I allocate to learning about how to do this in the best way
Anyway, in conclusion, the fact that there is no upper limit is great to know, thank you for all your help the last few days, it’s been much appreciated!
With multiple nodes. No one has made this with a single node. 
1 Like