June 14, 20242 yr So I just started using unraid with one Toshiba 8TB NAS drive and just recently bought one more and set it up to be my parity drive. The parity sync took 10 hours and completed with no errors. It's been something like 4~5 days and I just logged on in my server and got a warning about "Reallocated sector count". The new one is like: And this is the first one I started my array with when I didn't have any parity drives yet: I mean, I just bought the parity drive, is brand new as far as anyone would tell, right? Should I be concerned? They are the same model HD's, why the parity is the only one with that error? Random? Thanks in advance.
June 14, 20242 yr Solution 41 minutes ago, lelejau said: Should I be concerned? Depends. If the error count stays the same, then you are probably fine. If it keeps increasing, I'd return the disk for a new one. Don't manufacturer warranty return it if possible, the drives you get back from a manufacturer warranty swap are typically refurbished. Return it for credit and get a new one if possible. If the count stays at 11 and never increases, I'd keep the drive. Regardless, you always need to be alerted to changes in disk health, you shouldn't need to wait until you log back in. Make sure notifications are set up and working. You should be getting a daily "everything is OK" notification so you know the server can contact you with errors.
June 14, 20242 yr Author Thanks for the info about the notifications, just created a new server and using webhooks integration. Will keep an eye on that error count.
June 18, 20242 yr Author So today I got spammed with more warnings. That's it, driver is bad? I just opened a ticket in the store I bought. It's an expensive hard drive, can't accept a faulty one. In advance, how can I wipe it completely and safely before returning it to the store? Edited June 18, 20242 yr by lelejau
June 18, 20242 yr 36 minutes ago, lelejau said: how can I wipe it completely and safely before returning it to the store? Parity doesn't have a file system, or recognizable files. It's theoretically possible to run forensic data recovery software on it and possibly see fragments of content, but at a glance there would be nothing to even attempt to recover. If you really feel the need to remove any traces, you can unassign it from the array and and run a preclear on it, but unless you have a reason to believe you are targeted for an investigation I wouldn't bother.
June 19, 20242 yr Community Expert On 6/14/2024 at 4:43 PM, lelejau said: So I just started using unraid with one Toshiba 8TB NAS drive and just recently bought one more and set it up to be my parity drive. This is a special case. With parity plus one data drive, parity is a mirror image of the data drive. 12 hours ago, lelejau said: how can I wipe it completely and safely before returning it to the store? 11 hours ago, JonathanM said: ...you can unassign it from the array and and run a preclear on it, but unless you have a reason to believe you are targeted for an investigation I wouldn't bother. Do this!!!!
June 19, 20242 yr 3 minutes ago, Frank1940 said: This is a special case. With parity plus one data drive, parity is a mirror image of the data drive. Yep, I totally missed they only had one data drive. 12 hours ago, lelejau said: how can I wipe it completely Preclear will do it, just be sure you are operating on the correct drive.
June 20, 20242 yr Author I have unassigned my parity drive and already installed preclear app. Still can't see where I'm supposed to go to erase the disk... What am I missing?
June 20, 20242 yr Community Expert 8 minutes ago, lelejau said: I have unassigned my parity drive and already installed preclear app. Still can't see where I'm supposed to go to erase the disk... You should be asking this in the support thread fro the Plugin/Docker that you installed. For plugins, the link for the support thread can be found on the 'PLUGINS' tab. For Dockers, go to the 'DOCKER' tab and left-click on the Docker Icon. Then look for the link: EDIT: You do realize that preclearing a disk will write zeros to every bit on the entire disk. This effectively erases the disk. IF you are really paranoid, you can do it a total of three times which is the requirement for many US government agencies for disk disposal. Edited June 20, 20242 yr by Frank1940
June 20, 20242 yr Author Why is the speed so low? This drive should handle higher speeds, right? (TOSHIBA_HDWG480 N300)
June 20, 20242 yr Author tower-diagnostics-20240620-1135.zip Just one update: I had a bad ssh configuration in my network infrastructure that was causing it to be publicly accessible, just noticed today so many attempts to login with invalid passwords. Already disabled it. Edited June 20, 20242 yr by lelejau
June 20, 20242 yr Community Expert I didn't see anything. Lets see if there is a preclear log. Look here for that:
June 20, 20242 yr 42 minutes ago, lelejau said: Why is the speed so low? This drive should handle higher speeds, right? If it was healthy, yes. A failing drive can be painfully slow as the drive tries to access bad spots.
June 20, 20242 yr Community Expert It has only run 27 minutes. Let it run for a few hours and see what happens...
June 20, 20242 yr Author So it became faster after a while and it just finished with this error (discord notification): Despite the error, would you say its erased?
June 20, 20242 yr 1 hour ago, lelejau said: Despite the error, would you say its erased? Probably not, that's too fast for even a healthy drive, but it sounds like the drive has either died, or is close to dying. Can you get a SMART report?
June 20, 20242 yr Author 36 minutes ago, JonathanM said: Probably not, that's too fast for even a healthy drive, but it sounds like the drive has either died, or is close to dying. Can you get a SMART report? Extended or short ? --Edit I get a read failure when trying to do a short test. Edited June 20, 20242 yr by lelejau
June 20, 20242 yr Community Expert It should take about 16 hours for that part of the preclear operation for a good disk. (~2hours/TB) If the MBR is bad (and unreadable), it will be difficult to read anything on the disk. Granted that three-letter government agencies might get some bits-and-pieces off of it, but the cost in time and equipment required would be substantial. Unless you are a target of someone to whom money is no object, you should be able to safety return it on a RMA. (If you might be a target, then the last thing you should consider would be returning it!) Edited June 20, 20242 yr by Frank1940
June 20, 20242 yr Author 15 minutes ago, Frank1940 said: It should take about 16 hours for that part of the preclear operation for a good disk. (~2hours/TB) If the MBR is bad (and unreadable), it will be difficult to read anything on the disk. Granted that three-letter government agencies might get some bits-and-pieces off of it, but the cost in time and equipment required would be substantial. Unless you are a target of someone to whom money is no object, you should be able to safety return it on a RMA. ( If you might be a target, then the l;ast thing you should consider would be returning it I'm no targeted, just for privacy. I have no clue of what the RMA may do to the hard drive to try to deny warranty for me, I don't know. SMART test is failing, and the extended one spams me with tons of warnings while it's running. "Pending sector"... So, I pretty much think its dead? Damn. I just bought it, assigned as parity drive and let it build it. Should I have done it another way? I mean, I just can't believe me eyes. The HD just died with 1 week usage lol and as a parity drive.
June 20, 20242 yr Community Expert 10 minutes ago, lelejau said: So, I pretty much think its dead? Damn. I just bought it, assigned as parity drive and let it build it. Should I have done it another way? I mean, I just can't believe me eyes. The HD just died with 1 week usage lol and as a parity drive. As @Kilrah said, DOA hardware is a fact of life. Google bathtub curve reliability and you can read about it. It became a recognized phenomenon during WWII as large numbers of integrated complex electronic/electrical/mechanical systems were being deployed for the first time by the armed forces. (When I was a young engineer with Western Electric, we were told the story of an instance where after the battles in North Africa, some general decided that all of the vacuum tubes in the tank's radios should be replaced before the invasion of Sicily. The failure rate of those radios in the next couple of weeks was astronomical compared to radios with the old tubes!) That is one of the reason why many of us still do the preclear operation on our new HDs. Most drives (with infant mortality) will fail in the first hundred hours. With the massive size of today's HD's, that is about the time required for a single preclear cycle. I, personally, would rather have a failure during the preclear cycle than during a data (or parity) rebuild.
June 20, 20242 yr Author I see. If I have done preclear things would be much easier for me now. For example, now I wanted to completely erase the disk. So I started a preclear with operation "Clear Disk". If I bought a new one and want to certify that it is good before adding to my array, which operation should I choose? Verify disk? Seems like a reasonable choice, lol Edited June 20, 20242 yr by lelejau
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.