geekhack
geekhack Community => Other Geeky Stuff => Topic started by: suicidal_orange on Tue, 15 December 2020, 03:53:51
-
Booting took a long time today, sounded like the hard drive was doing something (yes, I still have a spinner :eek:)
This drive isn't even three years old (Christmas 2017) and has had an easy life but it's dead, right? SMART details below.
SMART Attributes Data Structure revision number: 10
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE
1 Raw_Read_Error_Rate 0x000f 081 065 006 Pre-fail Always - 129126238
3 Spin_Up_Time 0x0003 098 096 000 Pre-fail Always - 0
4 Start_Stop_Count 0x0032 094 094 020 Old_age Always - 7148
5 Reallocated_Sector_Ct 0x0033 100 100 010 Pre-fail Always - 0
7 Seek_Error_Rate 0x000f 066 060 045 Pre-fail Always - 4013813
9 Power_On_Hours 0x0032 096 096 000 Old_age Always - 3851
10 Spin_Retry_Count 0x0013 100 100 097 Pre-fail Always - 0
12 Power_Cycle_Count 0x0032 098 098 020 Old_age Always - 2223
183 Runtime_Bad_Block 0x0032 100 100 000 Old_age Always - 0
184 End-to-End_Error 0x0032 100 100 099 Old_age Always - 0
187 Reported_Uncorrect 0x0032 100 100 000 Old_age Always - 0
188 Command_Timeout 0x0032 100 100 000 Old_age Always - 0 0 0
189 High_Fly_Writes 0x003a 100 100 000 Old_age Always - 0
190 Airflow_Temperature_Cel 0x0022 083 056 040 Old_age Always - 17 (Min/Max 15/17)
193 Load_Cycle_Count 0x0032 097 097 000 Old_age Always - 7175
194 Temperature_Celsius 0x0022 017 010 000 Old_age Always - 17 (0 10 0 0 0)
195 Hardware_ECC_Recovered 0x001a 008 002 000 Old_age Always - 129126238
197 Current_Pending_Sector 0x0012 100 100 000 Old_age Always - 0
198 Offline_Uncorrectable 0x0010 100 100 000 Old_age Offline - 0
199 UDMA_CRC_Error_Count 0x003e 200 200 000 Old_age Always - 0
240 Head_Flying_Hours 0x0000 100 253 000 Old_age Offline - 1239h+15m+58.447s
241 Total_LBAs_Written 0x0000 100 253 000 Old_age Offline - 3855056683
242 Total_LBAs_Read 0x0000 100 253 000 Old_age Offline - 2429566323
SMART Error Log Version: 1
No Errors Logged
SMART Self-test log structure revision number 1
Num Test_Description Status Remaining LifeTime(hours) LBA_of_first_error
# 1 Short offline Completed without error 00% 371 -
# 2 Short offline Completed without error 00% 363 -
# 3 Extended offline Interrupted (host reset) 50% 311 -
-
Judging on the hours vs flying hours I'm guessing this is an external backup drive.
The ones to watch are:
Reallocated_Sector_Ct (IMPORTANT)
Reported_Uncorrected sector count (IMPORTANT)
Current_Pending_Sector (concern, as these are recoverable, though it may need a format to do so and failure to do so automatically is a bad sign)
Ignore the data, look at the raw values, everything else is interpretive and may be wrong unless you use the manufacturers software to read them, if these are ANYTHING but zero in raw data, get worried. In this case, all are zero, which is good, it means there was nothing to write and therefore not open to interpretation. No news is good news...
What is worrisome though is the error rates (seek, read and ECC recovered numbers).
These usually are not indicative of a failing platter but instead a bad cable, be it SATA or USB or a failing buffer. It's almost always the cable. The good thing with this type of error is can can monitor it after a cable change and see how your change effected it and see if things are better or not. Try a new cable and perform some large file transfers to and from using MD5 hashes to verify data integrity. Use large, freshly downloaded and verified files (linux ISOs work) starting from your computer to this drive, and back. If the MD5 checks out after a few large moves back and forth you're probably good to go.
Something to note. Even if you replace the cable and it checks out with md5 hash checks, the number may continue to rise fast on some of these because the data on the drive itself may already be compromised so those numbers are now questionable unless you copy everything off, format and put it all back then take note of the new numbers. This also won't fix any data corruption that has already happened but does clean up the errors in the files. Some files may even need to be opened and re-saved to correct them, missing data or not. If all this sounds like a pain in the neck, it is.
Remember none of this is 100%, it could fail tomorrow or last a darn long time, but in terms of what we can see it seems healthy at least firmware side.
If it was slow to boot and making any odd sounds, especially clicking, START BACKING UP NOW. These are signs of a mechanical failure in progress. Do not let it sleep, do not shut down, start backing up because the next time it power cycles or even accesses data may be the last time it ever reads that sector or spins up again. By the way, backup your windows key at the same time if you can but focus on personal data first then work out to less important stuff.
By the way under normal use desktop spinner drives have about a 50-75% failure rate by end of year 4 (Caviar Black is the exception, those get an extra 1-2 years), it being 3 years old is not a good sign. The numbers for failure start really high for year one, then drops really low for year two, then during year 3 they start ramping up. A 4 year old spinner is usually ready for retirement and if you find 7 or 8 year old one that has been in constant operation it's like meeting a 100 year old person.
Do not put any serious load, including a virus scan on a drive 4 or more years old without making a backup first, and note that backup process may kill it before the scan does. Doing a full backup or virus scan is the hardest a drive usually works for most people so it's a high risk operation. I've had numerous drives on customer systems fail this way.
-
As was mentioned, I would back the drive up immediately. I know I have a drive that is starting to go - currently it is pulled from the system so I can back it up once I get a larger drive (looking to move my array to all 4TB's from the current 10 - 2TB drives) and slowly start the migration of the array as I acquire drives to replace them. Fortunately there are good deals to be had on large capacity drives and normal capacity SSD's right now.
-
They've got these new QLC drives that are pretty reasonable in cost. for General OS use, I'd say they're fine.
You'd want something like the SN750 or EX950 for a midrange main drive.
There's always the 970evo and evo pro if you want luxury tax.
Backup first.
If you run chkdsk on boot, and it relocates ALOT of sectors, then the platter is rotting, usually it's moisture getting in.
-
Thanks for all the thoughts, liking the sound of a dodgy cable especially as there was no clicking, just spinning at different speeds (I've heard clicking and spin up to full speed -> stop [repeat] death throes before)
The drive is internal on my main PC and holds three copies of my music library (wav, flac and ogg) but most of my listening is during my commute these days so the drive is mostly idle which is why the hours are weird. No OS, no irreplaceable data, just many hours of ripping CDs if it dies. If the cable's corrupted any of the .wav the player will not handle it so I'll have to decompress all the flacs, that will get the error count up (or not if the cable was the problem)
-
They've got these new QLC drives that are pretty reasonable in cost. for General OS use, I'd say they're fine.
You'd want something like the SN750 or EX950 for a midrange main drive.
There's always the 970evo and evo pro if you want luxury tax.
Backup first.
If you run chkdsk on boot, and it relocates ALOT of sectors, then the platter is rotting, usually it's moisture getting in.
I snagged a 1TB inland Pro QLC drive for the wife's laptop and it has been great - the Samsung goodies stay in my rig and I tossed one of my many spare 500GB EVO's in my son's rig (have alot of 250 and 500GB Samsung SSD's).
Thanks for all the thoughts, liking the sound of a dodgy cable especially as there was no clicking, just spinning at different speeds (I've heard clicking and spin up to full speed -> stop [repeat] death throes before)
The drive is internal on my main PC and holds three copies of my music library (wav, flac and ogg) but most of my listening is during my commute these days so the drive is mostly idle which is why the hours are weird. No OS, no irreplaceable data, just many hours of ripping CDs if it dies. If the cable's corrupted any of the .wav the player will not handle it so I'll have to decompress all the flacs, that will get the error count up (or not if the cable was the problem)
I would swap the cable and the port that is being used on the Mobo and retest the drive and see if that clears up alot of the errors like LeslieAnn mentioned. If that doesn't improve then I would be looking to source a new drive - even if it is only media on it, it's still a pain to re-rip the media.
-
I had a similar warning on a 5-year old western digital black, and it died within months. make sure to backup anything important.
Later my HDD, CPU, Motherboard, and PSU all died at once on mini-itx build. I was using a silverstone SFX PSU which might have been why
-
Swapped the cable and port and no problem since so sticking my head in the sand.
my HDD, CPU, Motherboard, and PSU all died at once on mini-itx build. I was using a silverstone SFX PSU which might have been why
Really, everything? That must have been a really bad PSU and I though Silverstone were good for small ones. I'm probably due a major failure as next gen CPU won't be compatible with current mobos and I'm still on DDR3 hoping to skip to DDR5.
-
Really, everything? That must have been a really bad PSU and I though Silverstone were good for small ones. I'm probably due a major failure as next gen CPU won't be compatible with current mobos and I'm still on DDR3 hoping to skip to DDR5.
Yes... Someone at the store said something about mini-itx motherboards having inferior protection from PSU issues. Sadly, it looks like the tread is either ATX or mini-ITX. There is very little support for mATX with Ryzen.
I did build an mATX PC, but I had to get an AsRock motherboard. In the future I will just use ATX.
-
Yes... Someone at the store said something about mini-itx motherboards having inferior protection from PSU issues. Sadly, it looks like the tread is either ATX or mini-ITX. There is very little support for mATX with Ryzen.
I did build an mATX PC, but I had to get an AsRock motherboard. In the future I will just use ATX.
Hmm... there is less space for components but that seems a silly place to save money.
Just had a quick look and there are mATX Ryzen boards, just not X570 ones. My next build is planned to have a full coverage waterblock, all being well I have 2021 to build a case ready for the release of the Rembrandt APU in 2022. Don't think I've ever tried to plan anything this far in advance, what could possibly go wrong (https://cdn.geekhack.org/Smileys/solosmileys/laugh.gif)
-
Yes... Someone at the store said something about mini-itx motherboards having inferior protection from PSU issues. Sadly, it looks like the tread is either ATX or mini-ITX. There is very little support for mATX with Ryzen.
I did build an mATX PC, but I had to get an AsRock motherboard. In the future I will just use ATX.
Either that person was a doesn't know what they were talking about or you had a very low end ITX board.
Yes, they usually have fewer power banks but that means it can handle less power, it doesn't lead to stability issues unless you are overclocking an I7 or I9 and it certainly shouldn't fry the cpu (they're pretty bulletproof). Stability is not the same as flat out frying everything, even the highest end ATX board with far more banks of power will fail if the PSU fails. Also, unless it was a M.2 drive, your drive didn't even draw power through the motherboard. This was absolutely a PSU problem.
BTW, MATX is a dead/dying format, several manufacturers are dropping support for it entirely because no one's buying it.
Ryzen saw the loss first because it was a total socket redesign whereas Intel it was merely a minor refresh of older designs, hence cheap and easy to make an MATX board. When Intel finally hits 10 or 14nm expect MATX to disappear on them as well.
Side note.
Good ITX parts cost more, often 20-100% more than you would pay for the equivalent to what you would get for ATX. For example, a $100 ATX psu is going to be $120-150 for the same wattage and quality. It's also not always the same brands who offer the best parts. This leads people to buy what they think is a good part when in reality it's merely meh. It's becoming more standard but for the time being ITX is still a bit specialized for the better parts.
-
Yes... Someone at the store said something about mini-itx motherboards having inferior protection from PSU issues. Sadly, it looks like the tread is either ATX or mini-ITX. There is very little support for mATX with Ryzen.
I did build an mATX PC, but I had to get an AsRock motherboard. In the future I will just use ATX.
Either that person was a doesn't know what they were talking about or you had a very low end ITX board.
Yes, they usually have fewer power banks but that means it can handle less power, it doesn't lead to stability issues unless you are overclocking an I7 or I9 and it certainly shouldn't fry the cpu (they're pretty bulletproof). Stability is not the same as flat out frying everything, even the highest end ATX board with far more banks of power will fail if the PSU fails. Also, unless it was a M.2 drive, your drive didn't even draw power through the motherboard. This was absolutely a PSU problem.
BTW, MATX is a dead/dying format, several manufacturers are dropping support for it entirely because no one's buying it.
Ryzen saw the loss first because it was a total socket redesign whereas Intel it was merely a minor refresh of older designs, hence cheap and easy to make an MATX board. When Intel finally hits 10 or 14nm expect MATX to disappear on them as well.
Side note.
Good ITX parts cost more, often 20-100% more than you would pay for the equivalent to what you would get for ATX. For example, a $100 ATX psu is going to be $120-150 for the same wattage and quality. It's also not always the same brands who offer the best parts. This leads people to buy what they think is a good part when in reality it's merely meh. It's becoming more standard but for the time being ITX is still a bit specialized for the better parts.
well it was an Asus Mid-tier motherboard. Definitely not super low end from 2015. Either way, I will never get silverstone anything everagain
-
I will never get silverstone anything everagain
I put a Silverstone 450watt SFX in my Ncase when I got it (it was cheap and available unlike the Corsair), it was obvious on day one it wasn't going to cut it, too much heat in the box and too high power draw despite only a modest GPU.
I should have phrased what I said earlier a little better.
It's not that ITX stuff isn't good, it's just pushed harder/hotter and some stuff just can't cope so you have to next level all of it in many instances. Your PSU may have done fine in another box with more air, mine would have but in my little ITX box it was just getting baked to a crisp and hitting peak temps under even medium loads due to the heat coming off the GPU and CPU. Yours may have done the same. As for Silverstone, their psus aren't bad, but they also aren;t great, however few people actually need great. One time where you do need great is in a tiny toaster.
Hopefully it all got worked out.