One of my disks is dying and it’s looking bad for the node on it. I have another disk lying around, but it is barely big enough to fit the node on it.
If I get rid of the trash though, it would barely fit, and should then shrink even more little by little with time, to make some security room on its new drive (if enough data can be migrated from the failing disk, which isn’t certain at this point…).
I keep rebooting this Node (once a day approximately) because the disk fails and disappears from lsusb, with more and more data in lost+found…
And my stats are not looking good right now:
So I was wondering if I could just drop all data from the trash folder so I could attempt a migration of that Node right now and hope for the best… ‘coz waiting for the trash to naturally shrink does not feel like a safe idea right now ^^’
Any thoughts?
PS: I do realize it might be the end of the road for this Node but… I wanna try saving it just in case it’s not doomed yet ^^
The only risk you run into is that the satellite might call restore from trash. In that case your node might end up getting disqualified. That risk is relativ low but not zero ofc. Restore from trash has been called a few times so far so it could happen again.
But when restoring stuff from trash, do you restore ALL pieces from trash, or just a few pieces? If that’s the latter, surely that wouldn’t be enough to DQ a node, would it?
Anyway I tried rsyncing my node’s data to another disk, and that finished killing the source disk after only 50GB transferred, I think. It does not appear anymore on the OS, and does some weird rattling and metallic clic sounds now…
So I think it’s definitely dead. RIP my 5TB node which is gonna get disqualified in a couple of weeks.
It had a good life, although the disk died after 2 years only which is not a great score ^^’
First loss after almost 4 years of SNOing
It had to happen eventually
Sure, it was a second-hand SEAGATE Barracuda Compute 8TB drive (SMR technology) that I got 2 years back (mid-2021) from a friend who had bought it in the beginning of 2021 and barely used it.
I’m surprised to see that the Date Of Manufacture is Sep. 2018, I’m wondering what happened to this disk between 2018 and 2021… I didn’t check any of that (sticker, smart info…) when I got it back in mid-2021 I must say.
That’s a must for any FIRST/second/nth hand. I have come across many that had Barracuda as their failed disk. Storj has made me opt for Enterprise grade and I love it otherwise I go for SSD.
Storj is full of awesome people that help you learn. Let that be Awesome Alexey or other mods that help you with your issue or fellow SNOs with their personal expertise. Don’t forget devs that take time from their sprints to educate us. Every day is a good day to learn.
I understand if HDD optimized for different workload performs poorly under the other workload. Like hosting databases on SMR drive.
But I never accept HDD dying under that different workload. Outside of random one-offs, this is indicative of garbage manufacturing ethics, vendor greed, and otherwise disregard of consumer. Modern technology is fully capable in throttling the load to prevent damage. Ah, you say they skimped on ball bearings and used crappy plastic for the arm coil — well, that’s shitty and such products should not be purchased, up to boycotting the whole vendor. Vote with your wallet.
Funny, how i tought while reading without reading the autor of the post sounds like arrogantrabbit wrote this
then
Congratulations, @arrogantrabbit your style is remarkable.
Even if our Opinions are sometimes far away from each other, i have to say, you are right.
If one of my hdd/ssd would die shortly after warranty is over, id never buy again from this manufacturer. I lost trust in Adata eg.
I dont like seagate, i had to much troubles with them. 2 days ago died in one moment another seagate hdd. It has warrenty, but to send it to seagate in holland is 30 euro, is a half of it price of new one. I made some troubleshooting, it is shortsercut in controller board. one place heating to 62 degree, but there is no components on this place.
I cant complain about seagate I only had issues with SMR drives, Ive had more hds die from other manufactures then I had die with seagate. Most of my drives that I have that still work are iron wolf pros. And a few 1TB drives from over 12 years ago that still work fine.
RIP my 5TB node 
which is gonna get disqualified in a couple of weeks.
sounds like arrogantrabbit wrote this 