Hello!
Got a several nodes with several questions:
Hello
It depends: if the different location is a near place, maybe the easiest way is to just move the HDD to that location. If it is not, I think the recommended way is to use rsync, which allows you to move the data with the service running most of the time - check the link below. However, if it is through the network it will take time depending on the HDD size.
I’m currently doing that using rsync from WSL2 in Windows to copy the data to a Rpi through the network (currently 24h for 1.5 Tb - but it depends on multiple factors like HDD and network performance)
This is not feasible without a solution like RAID or similar (or something like mergerfs). The easiest way maybe it´s just running a second node on the extra space of the second disk and complete the vetting process.
About the rsync process, you can find more information here
yep, thaks alot!
in p.3 could establish RAID (got a technical ability with no skills in that).
is the max capacity of the node still about 20TB or so?
Thank you for an answer Alexey!
There is no max at all.
Hi!
Got another question about this case:
I need to merge 2 different full nodes (email; wallet; identity - are differs) in one place with same net and provider settings.
For now it works in different isolated from each other networks, the 1st one with static IP, the 2nd goes with dynamic IP through No-IP service.
Is there could be some penalties for that?
I mean disqualification, less traffic, some net incompatibility cause of different settings, etc…
I’ve did that about a couple of years ago in a beginning of storj node operating cause of restrictions of same account settings in that time.
You can’t merge nodes together, but you could both nodes on the same machine from the same IP, storing data on different hard disks. Running them from behind the same IP means they will be considered as one node when being selected for uploads (ingress), effectively halving your potential ingress per node. Since your nodes are already full it means you won’t really see any difference, provided your internet bandwidth is sufficient.
Not sure that I mean what I have described.
To clear my question: is there some penalties for two different nodes where wallet/email/identity are differ on both nodes which gets bandwidth from the same IP?
And considering for manual the minimal bandwidth for one node is 25mb/sec, so if I have got 100mb/sec I can elevate not more than 4 nodes, correct?
Well, the terms of service state that you should use the same ERC-20 compatible address for all of your nodes. But as far as I know there is no penalty with respect to how your nodes work if you go against that. I’m sure Storj could enforce this more strictly in the future if they wanted to.
As far as bandwidth is concerned, you would have to evaluate that yourself. It is possible you could run more than 4 nodes on 100 Mbps, but if that is not symmetrical (i.e. you have 100 down/10 up) than your upload speed could be your bottleneck. There are too many factors to say how many nodes any one connection will support.
I have got 100 up/down band with fiberoptics in last mile from provider server,
so for now I have already had proper preparations for that.
And got an ability to expand to 1gbps with ease, it just cost alot and my needs a far from that band.
The capacity of HDDs 12TB/each.
In summary of that can I expand to 4 nodes or I can elevate more?
If you hardware is up to the task, I see no reason you couldn’t expand beyond 4 nodes. If they are all USB connected drives you could run into a bottleneck there. Or if you are using a low power device (Rpi or similar) you might be limited by CPU/ram. I expect 4 would be near the upper limit if running on an Rpi 4 for example.
Not a good idea to break ToS unless Storj Labs has specifically stated it’s allowed. In this case there is definitely a penalty if you don’t reach the minimum payout threshold on individual addresses. You much more likely would if you have all set to the same address.
Not really, because they would share traffic on the same IP. You can easily run 10 nodes on an RPi 4 from one IP.
Is identity not links to a certain wallet and email address that put in the beginning of a node creation?
There is no goal to save the current statement, I have rolls up a several nodes about one and a half year ago in just curiosity in a proccess, so if there is a common practice to change a wallet in a node Docker startup command in Power Shell without penalties, node crashing and some other time and nerves pain I will change it gladly.
Identities need to be separate, but you can use the same wallet and email address on all nodes. In fact the ToS say you HAVE TO use the same wallet address on all nodes you manage.
Is there some specific to change wallet address or just need to roll up the Docker parameter string with the same '"-e WALLET=" on all nodes?
That’s all there is to it. You do have to first remove the container before doing the run command again. But then you’re all set.
Ok, thanks for an answers!
And good day!
Seems like that my this year epic journey with a new server room installation comes to an end and I’m on rolling up the Debian 11 to my custom build server.
After looking to a system requirements I’ve got a question - is Debian 11 suits for rolling up the node? In a table of consistency just only 9 and 10 versions.
My nodes work just fine on Debian 11.