Yes, but that link explains nothing. Does the storage folder need to be in an elevated privilege (must be root) directory as I’ve seen it stated elsewhere in this forum? Should it be a subdirectory of something? Here’s the crazy place I put my first node:
–mount type=bind,source="/home/thaistar/.local/share/storj/identity/storagenode",
destination=/app/identity
–mount type=bind,source="/home/thaistar/.local/share/storj/",destination=/app/config
The storagenode data in general should be placed on a separate HDD. You can select folder or root of the drive, it’s up on you.
If it’s the system drive, you can select almost any folder would you like.
The user from whom you start the container must have read and write permissions, for simplicity you can provide full rights.
If you run a container with sudo, then the owner will be root. If you decide to run container without sudo later, you should grant permissions recursively to your user.
I see two choices. Run a node software for each HDD, would probably require multiple identities I think but you can register under the same email now (Not sure about the identities part).
Second choice is just setup a mirrored storage pool, it’s what I have done for my own data. Easier, less complex and faster setup if you ask me.
If your interested, I can help ya set it up, contact me on discord @ SpyShadow#8144
Hello @SpyShadow,
Welcome to the forum!
Each node must have an own unique signed identity. So you need an own identity and own token.
Hi guys,
So much interesting discussion!
I am considering the “1 disk per node” option. But I would like your opinion on the best way to proceed.
My current configuration:
I am running several virtual machines on a personal ESX server.
My current node is on a debian VM, configured with 1 vCPU and 1 GB RAM.
My question
Assuming I will create a new node with a new disk of 3 TB, should I:
- create a new VM (with same configuration) => this option would need to allocate 2 vCPU and 2 GB RAm for the total of my 2 nodes.
- create a new docker container on the existing image => this option would allow to reduce the amount of resources (maybe no more resources or maybe just some more RAM? I don’t know).
I would like to have the best option to limit the resources use without impacting performances.
Thanks for your opinion!
I would recommend to increase resources for the existing VM to handle two nodes, i.e. 2 vCPU and 2GB of RAM and run the second node there.
In any case your second node should use a different physical disk for storage.
If you want to use the same disk, then better to increase the allocation for the first node and do not run the second.
4 Likes
Thanks.
Is it really mandatory to double ram and CPU?
On the vSphere client dashboard, I see that my Storj VM uses only few percents of the allocated vCPU, as you can see below (orange curve, measured by the right scale):
Does anyone have some experience on that?
Thanks
It’s not mandatory, but at least a CPU in the requirements: https://documentation.storj.io/before-you-begin/prerequisites
Regarding RAM - you can try to use the current 1GB, but your nodes could perform worse than a one node.
1 Like