Goodbye and Goodluck

Uh no I would not sell this hard drive to anyone why would I want to put someone else though the same pain that I have been though with it?

4 Likes

you can add info about it, then it will be choice of byer. also you can make some support ticket to manufacture, may be it is known problem need just fw update

1 Like

On the flip side, I’ve had more CMR drives fail silently than SMR, with ZFS showing constant errors but SMART insisting it was fine. Hell, I still use SMR in my off-site backups and those drives have been going strong for six years now.

SMR shouldn’t be used in a NAS or anywhere speed is of the essence and there’s been some hella shady marketing around them, but it’s pretty universally agreed online that SMR drives are no less reliable than CMR in terms of actual failure/error rate. You almost certainly just have a bad drive.

Also, the system handles spin-down on its own (TrueNAS etc has a setting that has it disabled by default, but can be set to normal per-drive - non-NAS OSes will jsut do it automatically), it’s not something you’ll be doing manually if you’re using it as a desktop drive.

If you use it for cold storage it may work but if your using for storage all the time and you have a power outage you can be sure your losing data. Ive had these drives for less then 6 years and they have lost data randomly over those 4 years that I had them running as storagenodes and also didnt DQ me but out of 2 of the 3 I had either lost my identify files and the last one DQed my node all together because I ran gradeful exit on them and it had failed. I guess depending on the brand maybe one SMR drive is better then the other but Ive had terrible luck with seagates SMR drives. On that note I have removed all SMR drives that ive had from my desktop and storagenodes.

Many people do this with external backup drives. Turn it on, backup and eject. I was just saying this is a bad idea for SMR drives.

I do, it’s just that I have not seen any reports anywhere that say SMR drives carry the risk of data loss. And I’ve looked for them. If this were a real risk, it would be all over the web. I’m not sure what happened in your scenario, maybe the HDD is just faulty in some way. Maybe it’s just a file system issues, which are more likely on SMR drives because they spend much more time in the writing state. But no, your anecdotal evidence combined with the lack of any reliable reporting on this is not enough to convince me there is an inherent issue with SMR drives themselves.

Going by your last post it does seem that you were using them in circumstances that are not ideal for SMR. And if power outages were a factor, file system issues are a fairly likely cause of loss as well. The large cache these drives have works against them in power loss scenarios.

I wish I keep track of all the weird things ive had with my SMR drives the recent one is the fact I used it as storage for my VMs pfsense to be exact, I turned system off gracefully and I went to boot it up and the boot was missing for my pfsense. I upgrade my backup power to try to steer away from the weird things that have happened but it seems to keep following me no matter what, And the drive I was using is directly from my desktop cause it was to slow for anything.

If you can find firmware for a seagate ST4000DM004 by all means let me know if there is a fix for these…

Have you searched here :point_down: ?

https://apps1.seagate.com/downloads/request.html

It requires serial number so I can’t search it for you.

Yeah I did nothing for my drive…

what firmware version do you have?

That I do not know, I have it in my server and proxmox doesnt show firmware versions
Nevermind it shows some info

Seagate ST4000DM004 - HDD FAQs
check this, looks like you have bad hdd

Logical Failure Examples:
Accidental deletion, accidental format, file corruption, software bugs, file system corruption, viruses and malware, and many, many

There is also a utility to download firmware

https://www.seagate.com/support/kb/firmware-update-utility-instructions-and-faq-004559en/

Hmm that sounds exactly the problem an i have 4 of these drives…

They are intended as a source of Neodimium magnets for you DIY garage. :grin:

3 Likes

@digitalfrank is there any trick on that? I’m running a dedicated-server in data-warehouse with 1 Gbit Netlink and 50 GB storage (currently 30 GB linked to STORJ-Node). Only 6 GB is used! The node has been running now for month.

I’m not expecting to get rich here, but STORJ rewards should at least cover hardware/server/running costs! At the moment I’m getting 5~8 USD per Month. It would not even cover electricity costs if I would run this “at home” and it’s far away from covering Server costs at all.

Any tipps or tricks?!

I hope you meant TB for all of those GB’s there. :smiley:

I have a tip: Readjust your expectations. It will take time for your node to collect data and start earning. @digitalfrank must have multiple nodes in different locations/IP ranges, otherwise he couldn’t have collected that much data. Take a look at this estimator to get an idea of what is realistic to expect per /24 IP range. Realistic earnings estimator

2 Likes

14 posts were split to a new topic: ZFS optimizations for usage with storagenode

I have 22 nodes in 10 different subnet /24 . and i have 0 costs of elettricity because i have 3 kw of solar panel.
I have 2 server proliant dl380 g8 e 3 storageworks d2600, in the next mounth i will activate others 10 nodes with 5 others different IP. I have 2 internet connection fiber 10gbit/2gbit and failover in 5g. ;-))

1 Like

Be careful with these upgrades! If Vadim senses you geting close to 200TB, he will double his nodes. :grimacing:
Just do it Ninja style :ninja:t4:.

4 Likes