My SSD now has 226 bad sectors. Is that significant?

Question

Ubuntu 13.04. The number of bad sectors on my SSD has steadily increased until 226 bad sectors today.

The thing is, I don't know whether 226 bad sectors is 0.0001% of the SSD, 1% of the SSD, or 99% of the SSD.

I haven't found anything on the internet this morning to answer this question.

The Disks utility doesn't tell me either, and I can't find the information in SMART Data.

---

sudo smartctl -a /dev/sda shows this:

smartctl 5.43 2012-06-30 r3573 [x86_64-linux-3.8.0-31-generic] (local build)
Copyright (C) 2002-12 by Bruce Allen, http://smartmontools.sourceforge.net

=== START OF INFORMATION SECTION ===
Device Model:     Samsung SSD 840 PRO Series
Serial Number:    S1ATNEAD645474H
LU WWN Device Id: 5 002538 5503c15c0
Firmware Version: DXM05B0Q
User Capacity:    256,060,514,304 bytes [256 GB]
Sector Size:      512 bytes logical/physical
Device is:        Not in smartctl database [for details use: -P showall]
ATA Version is:   8
ATA Standard is:  ATA-8-ACS revision 4c
Local Time is:    Fri Jan 24 20:37:08 2014 CST
SMART support is: Available - device has SMART capability.
SMART support is: Enabled

=== START OF READ SMART DATA SECTION ===
SMART overall-health self-assessment test result: PASSED

General SMART Values:
Offline data collection status:  (0x00) Offline data collection activity
                                        was never started.
                                        Auto Offline Data Collection: Disabled.
Self-test execution status:      (   0) The previous self-test routine completed
                                        without error or no self-test has ever
                                        been run.
Total time to complete Offline
data collection:                (53956) seconds.
Offline data collection
capabilities:                    (0x53) SMART execute Offline immediate.
                                        Auto Offline data collection on/off support.
                                        Suspend Offline collection upon new
                                        command.
                                        No Offline surface scan supported.
                                        Self-test supported.
                                        No Conveyance Self-test supported.
                                        Selective Self-test supported.
SMART capabilities:            (0x0003) Saves SMART data before entering
                                        power-saving mode.
                                        Supports SMART auto save timer.
Error logging capability:        (0x01) Error logging supported.
                                        General Purpose Logging supported.
Short self-test routine
recommended polling time:        (   2) minutes.
Extended self-test routine
recommended polling time:        (  20) minutes.
SCT capabilities:              (0x003d) SCT Status supported.
                                        SCT Error Recovery Control supported.
                                        SCT Feature Control supported.
                                        SCT Data Table supported.

SMART Attributes Data Structure revision number: 1
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE
  5 Reallocated_Sector_Ct   0x0033   097   097   010    Pre-fail  Always       -       238
  9 Power_On_Hours          0x0032   099   099   000    Old_age   Always       -       1331
 12 Power_Cycle_Count       0x0032   099   099   000    Old_age   Always       -       341
177 Wear_Leveling_Count     0x0013   097   097   000    Pre-fail  Always       -       75
179 Used_Rsvd_Blk_Cnt_Tot   0x0013   097   097   010    Pre-fail  Always       -       238
181 Program_Fail_Cnt_Total  0x0032   097   097   010    Old_age   Always       -       238
182 Erase_Fail_Count_Total  0x0032   100   100   010    Old_age   Always       -       0
183 Runtime_Bad_Block       0x0013   097   097   010    Pre-fail  Always       -       238
187 Reported_Uncorrect      0x0032   100   100   000    Old_age   Always       -       0
190 Airflow_Temperature_Cel 0x0032   067   062   000    Old_age   Always       -       33
195 Hardware_ECC_Recovered  0x001a   200   200   000    Old_age   Always       -       0
199 UDMA_CRC_Error_Count    0x003e   099   099   000    Old_age   Always       -       1
235 Unknown_Attribute       0x0012   099   099   000    Old_age   Always       -       128
241 Total_LBAs_Written      0x0032   099   099   000    Old_age   Always       -       33308592070

SMART Error Log Version: 1
No Errors Logged

SMART Self-test log structure revision number 1
No self-tests have been logged.  [To run self-tests, use: smartctl -t]


SMART Selective self-test log data structure revision number 1
 SPAN  MIN_LBA  MAX_LBA  CURRENT_TEST_STATUS
    1        0        0  Not_testing
    2        0        0  Not_testing
    3        0        0  Not_testing
    4        0        0  Not_testing
    5        0        0  Not_testing
  255        0    65535  Read_scanning was never started
Selective self-test flags (0x0):
  After scanning selected spans, do NOT read-scan remainder of disk.
If Selective self-test is pending on power-up, resume after 0 minute delay.

Solution 1

It eventuated that that Samsung 840 Pro SSD had a catastrophic failure later in 2014. It was replaced under warranty with the Samsung 850 Pro, which has now been running successfully since then, and still does not report any bad sectors. I'm pleased about that, because I thought maybe it was an "undocumented feature" of SSDs.

Solution 2

The smart data of our SSD says that you used the SSD for 1331 hours and your SSD has suffered 33308592070*512/1024^4=15,5 TiB writes. That means ca. 12 GiB writes per power on hours on your SSD. That is kinda lot for a typical consumer SSD usage. What do you use the SSD for?

You are still way below the write cycle limit of your SSD, which is an MLC based one, so should tolerate ca. 3000 write cycles, which would mean ca. 256GB*3000=768 TB writes. So I would say you are safe.

But your SMART data shows that now you have already 238 reallocated (failed) sectors. That number is still very low compared to the few hundred million sectors on your drive, but for me it is surprising that after only 60-70 write cycles your SSD has failing blocks. Do you have some spare space on your SSD? The SSDs needs some spare space to be able to efficiently manage the wear on the drive. If the drive is almost full it wears faster because of the increased writes of the wear leveling algorithm.

Here is a graph showing a 840 PRO 256 GB SSD under write test. It has still zero, or very close to zero reallocated sectors after 300 TB writes. You have more than 200 after less than 20 TB.

You are using the latest firmware of the drive. I don't want to worry you but I think you should contact Samsung with that SMART output, and ask about their opinion. I think your drive has some kind of problem.

---

UPDATE:

After Samsung's response - who basically said everything is fine - I would say that as long as you don't encounter data loss, (that is as long as the badblocks discovered during a write operation) I think you shouldn't worry then.

You can see the badblock count in the Runtime_bad_block line (238, the Raw value) that is the sum of the Program_Fail_Cnt_Total (238, which is the failed write operations) the Erase_Fail_Count_Total (0, which is the failed erase operations) and the failed read operations. So the failed read operations on your drive at the time the smart output was generated is 238-238-0=0, so you didn't have any failed read, you don't have any data loss.

If your drive starts to encounter failed reads too in the near future (that means the Runtime_bad_block will not equal Program_Fail_Cnt_Total + Erase_Fail_Count_Total), potential data loss, I would recontact Samsung. Until that time just enjoy your SSD.

Comments

[mcantsin]

Did you check this? Try sudo badblocks -v /dev/sda in the command line

[el_gallo_azul]

Disks utility reports 227 bad sectors this afternoon (after about 8 hours since I posted this question), and the accompanying SMART Data says 183 runtime-bad-block-total 227 OK. Knowing that, it's not so bad since 227/250059095=0.000000908 | In reply to the question from @mcantsin: sudo badblocks -v /dev/sda gives Checking blocks 0 to 250059095 Checking for bad blocks (read-only test): done Pass completed, 0 bad blocks found. (0/0/0 errors) which hopefully is a good thing.

[falconer]

@user91583 Please add the output of sudo smartctl -a /dev/sda to your question. (Or if your SSD is not /dev/sda then change that command accordingly.)

[falconer]

@el_gallo_azul Then first install it by running sudo apt-get install smartmontools .

[el_gallo_azul]

@falconer sudo smartctl -a /dev/sda now gives pastebin.com/vGrQMz1T (I hope this link works for you. Let me know if it doesn't).

[Jason C]

I know this was a while ago, but just out of curiosity; how did you set up that drive to begin with? Was it cloned from another drive? Also, what filesystem(s) are you using?

[el_gallo_azul]

I bought the computer with Ubuntu 13.04 preloaded onto the SSD, and also with an internal HDD. I started this post when I was still using Ubuntu 13.04. I don't know how Ubuntu 13.04 was installed by the vendor. Since then: my motherboard failed, it was repaired, I updated to Ubuntu 14.04, I ended up with a lot of problems, I did a clean reinstallation from scratch of Ubuntu 14.04, got everything running again, and since that date, according to the Disks utility, the number of bad blocks has not increased from 302. The SSD and the HDD use Ext4 filesystem.

[kagali-san]

I wonder if it's much relevant, but I just had to purge 1024 512-byte sectors with hdparm (--write-sector) from my TOSHIBA THNSNH256GMCT. Probably will demand explanations on SU later :)

[el_gallo_azul]

Thanks. It's no longer relevant to me. After my motherboard failure in February 2014, I had installed Ubuntu 14.04 and things were going OK. But then I had a sudden, catastrophic SSD failure in September 2014. The failed Samsung 840Pro SSD was replaced under warranty with a Samsung 850Pro SSD, and I successfully reduced the number of daily write cycles to the SSD. So far, so good.

[el_gallo_azul]

I use BOINC and I also let the SkyNet Java client run in multiple instances in the background. Yes I let it do very much writing in this process, and I am trying to find a way to get both of these to use the HDD instead of the SSD. On the spare space matter, my SSD is 25% used, and I will limit its use to 70% if I can - so far, so good. That's interesting about the performance of my SSD against Samsung benchmarks. I will follow that one up if I am able.

[falconer]

@el_gallo_azul I found an updated version of that test which I linked in my answer, here it is. It says explicitly the numbers after 500 TB: The 840 Pro, which has recorded only two reallocated sectors and no unrecoverable errors, has the same "good" health rating. So after 500 TB writes the same model as yours had only 2 reallocated sectors, you have more than a hundred times of that after less than 20TB. I would definitely send an email to Samsung with your SMART data and ask their opinion. It won't hurt.

[el_gallo_azul]

I just sent the information via webform to Samsung. I intend to post a summary of the Samsung response here if I receive one. Thanks for your help.

[el_gallo_azul]

First response from Samsung on 29.01.2014: "You want to know why would you have many times more reallocated sectors in your Samsung 256GB SSD PRO 840 Series. Please be advised that we forwarded this concern to the relevant department and we will contact you once update becomes available."

[falconer]

@el_gallo_azul Thanks for keeping us updated. I'm interested in Samsung's response.

[el_gallo_azul]

Response from Samsung received today 31.01.2014: "This is in relation to your email on 28th January 2014 regarding your enquiry on our Samsung Solid State Drives (SSD). I have received the feedback from our Level 2 support and he advised that each models results will vary depending on the condition they are to be used in and for period of time they are used. Those results that you see are essentially used in a control environment not as everyday use. We can''t say that every single SSD will have the same results so it may not have the same reallocated sectors as the one before."

[el_gallo_azul]

So essentially I interpret their response as "That's acceptable". I'll continue to find and apply ways to reduce the number of write cycles.

[falconer]

@el_gallo_azul Not the response what I expected, but it looks like Samsung thinks it is fine (and they must know better then me :D ).

[el_gallo_azul]

Up to 296 bad sectors today according to the Disks utility. I have 5 BOINC tasks left with a couple of days left to go. Once they finish, I'll move the BOINC working directory from the SSD to the HDD and (hopefully) that will sort most of the problem out.

[el_gallo_azul]

I was over 300 bad sectors yesterday. My computer then bit the dust yesterday, so I can't do anything about it until the thing is up and running again. Shame. I have to use my previous old clanger in the meantime - lucky I still have it.