It just uses the default location there, but nowhere does it say it HAS to be there. And in the docker run command you get the option to point to it, which makes it pretty clear you can put it wherever.
You’d probably get that because both of those would have given you your answers. But in my experience that reaction is actually pretty rare on the Storj forums. Especially if you’re asking input about a specific situation.
You’re saying you don’t see people saying what you should have done differently. But I’ve seen several people telling you you shouldn’t use NFS as it’s not supported. I see people offering tips to prevent the node from starting if the data location isn’t available by using a subfolder or putting the identity on the same HDD. I’ve suggested myself that after a migration you should check the logs to make sure nothing is going wrong. It’s been explained to you why you can’t go back to the old node even if you’re not yet disqualified. I think you’re just angry about what happened, which is understandable, but you’ve been getting the tips to do this right in the future. People have been helping you.
And yes you can now run more than one node per IP. They would simply share incoming traffic if you do. So the same traffic split over 2 nodes instead of one.
So how would I do that? just run multiple docker instances with different mount points? or is there actually any documentation on how to do this? Last time I set up a node from scratch was when they first stared v3.
And yes I am upset and ran in here a little hot. To be fair I had only been awake about five min and the first thing I did this morning was check my node since I had just done the migration. Even if it was not working these results are not at all what I was expecting.
Ok so I found some documentation about multiple nodes and what I am gathering from it is that it is not recommended. Which is why I don’t think I ever revisited the idea. But now I am thinking it may not be such a bad idea because then it would not be putting all of your eggs in one basket so to say. Like I did and just paid for it.
We get that but you have been professional in raising your issue and at least didn’t resort to calling Storj - scam or cheaters; which we have seen a few times.
I would strongly recommend to get email notifications about announcements from forum or better yet just make a post and ask a question about something you want to do. We will always help as a good community.
I also lost 4-5 nodes, becase some problems, it is normal. Now i am just more careful about making some things. Own experience is the best teacher, take very expencive price, but make all clear.
I ran 21 nodes on 4 PC, All windows GUI nodes. I dont like docker, because i do not know how it work.
If you’re a linux admin, then you should have come across this problem with other stuff too. For example, when do automated backups to a different drive, then you risk filling your root filesystem if you write directly to the mount point if there is a problem with mounting the drive. It’s just best practice to never write directly to a mount point.
I agree that it should be written in the docs and several users including me asked numerous times for it to be added to it. I wouldn’t expect the average user to know about this problem, but a linux admin should definitely know better.