(i edit my original post)
(running windows 10)
so i have the Node working 100% now,
i have a tinny single 2TB disk running at the moment,
now i have a pcie raid card and a whole bunch of hard drives i want to configure on raid 5.
how do i go moving existing data on my D: drive (strorj drive) into the new raid 5 storage? is it as simple copy and paste and assign the D drive letter?
It’s not recommended to use RAID, especially RAID5:
With todays disks it’s easy to lose a whole RAID volume during the rebuild because of bitrot.
Please, avoid this problem by running an own node on each disk separately. Since all nodes behind the same /24 subnet of public IPs will share the ingress, so they would act as a RAID on a network level.
If the disk would fail - you will lose only one node, but others will remain, so actually you will lose only part of the data. In case of RAID you likely will lose everything.
You of course can configure a RAID10, it will be more robust, but it will be twice as expensive.
im currently a miner and i have tons of solar, i work on IT, so i get a bunch of free hardware some of it hardly used… i put up this node at almost no cost. everything is on a big UPS. (bought a sata pcie card for this project)
i have about 20TB+ worth of hard drives around the shop and why not put them to a good use.
with the sata pcie card i was thinking adding more hard drives to storj but for what i read storj can only point to 1 single drive?
i could make a huge virtualized hard drive?
make the drives Raid 1?
i could run Hyper-V and have a few operating systems, maybe that way?
Check out snapraid. Its raid without the risk of losing everything. Its just parity, so you can rebuild, but if that fails you only lose the data that was on the lost disk. Its not calculate on write so youll have to decide how often to sync the parity.
I would not recommend such “RAID” where parity is not calculated on write. This is the same as a spanned drive then.
In case of storagenode it will mean only one - it will be disqualified.
So, it’s better to have independent nodes.
Well I’d say STORJ network usage is the limit.
Currently, 20TB would take years (6+ currently) to fill up unless you have a way to run nodes on different /24 subnets.
So maybe start small, and ramp up progressively as your nodes fill up one after another.
Good initiative, but be aware that it doesn’t stay stable with time as the more data you have, the more deletions occur which results in less “net-ingress” overall.
You might wanna check this nice community made estimator to have a better idea on what to expect: