Community
cancel
Showing results for 
Search instead for 
Did you mean: 
idata
Community Manager
1,636 Views

Unrealistic SMART values

Hi there,

I bought my Intel 320 SSD in January and use it as my system drive. So, not much writing, I just boot from it. My data is stored elsewhere.

Nevertheless I get these terrifying SMART values:

241 Host_Writes_32MiB 0x0032 100 100 000 Old_age Always - 48989

242 Host_Reads_32MiB 0x0032 100 100 000 Old_age Always - 19852

Does this tell me that 1.53 TiB were written on this drive?! Because that's pretty much impossible. I rarely write anything on it, so the Reads value should be higher than the Writes value (they're both too high, though).

Here are the other values, some of which I'm unable to interpret:

3 Spin_Up_Time 0x0020 100 100 000 Old_age Offline - 0

4 Start_Stop_Count 0x0030 100 100 000 Old_age Offline - 0

5 Reallocated_Sector_Ct 0x0032 100 100 000 Old_age Always - 0

9 Power_On_Hours 0x0032 100 100 000 Old_age Always - 1591

12 Power_Cycle_Count 0x0032 100 100 000 Old_age Always - 305

170 Reserve_Block_Count 0x0033 100 100 010 Pre-fail Always - 0

171 Program_Fail_Count 0x0032 100 100 000 Old_age Always - 0

172 Erase_Fail_Count 0x0032 100 100 000 Old_age Always - 0

183 Runtime_Bad_Block 0x0030 100 100 000 Old_age Offline - 0

184 End-to-End_Error 0x0032 100 100 090 Old_age Always - 0

187 Reported_Uncorrect 0x0032 100 100 000 Old_age Always - 0

192 Unsafe_Shutdown_Count 0x0032 100 100 000 Old_age Always - 23

199 UDMA_CRC_Error_Count 0x0030 100 100 000 Old_age Offline - 0

225 Host_Writes_32MiB 0x0032 100 100 000 Old_age Always - 48989

226 Workld_Media_Wear_Indic 0x0032 100 100 000 Old_age Always - 34264840

227 Workld_Host_Reads_Perc 0x0032 100 100 000 Old_age Always - 1

228 Workload_Minutes 0x0032 100 100 000 Old_age Always - 2366980413

232 Available_Reservd_Space 0x0033 100 100 010 Pre-fail Always - 0

233 Media_Wearout_Indicator 0x0032 100 100 000 Old_age Always - 0

What's going on here?

Tags (1)
0 Kudos
11 Replies
CFran6
Novice
155 Views

Intel 320 SSD spec from March 2011 specifies the units of attribute 241 and 242 as "LBAs" but attribute 225 as "32 MiB". This does not make much sense because the raw value of 225 and 241 are apparently always identical. Smartctl test outputs we received on smartmontools-database mailing list suggested that 32 MiB is likely the more reasonable unit. Therefore I decided to use this for the attribute name in the smartmontools drive database. "LBAs" from the spec does not make any sense for your device as 19852 read sectors (less than 10MiB) are probably not enough for a single boot :-)

This device should also support the new ATA "Device Statistics" feature which is (unlike SMART Attributes) part of the ATA Standard. Please try "smartctl -l devstat ..." and report the result here, in particular the values of "Logical Sectors Read" and "... Written" from "General Statistics" log page. This should give a hint which unit of measure is actually used for the related SMART Attributes.

BTW: The "Workload" related attributes 226-228 need to be cleared before being useful. This could be done with "smartctl -t vendor,0x40 ...", see smartctl man page.

Thanks,

Christian

(smartmontools.org)

idata
Community Manager
155 Views

Hello Christian,

Thank you for your reply, it's nice to instantly get feedback from a smartmontools dev on this issue.

Here's the output:

Device Statistics (GP Log 0x04)

Page Offset Size Value Description

1 ===== = = == General Statistics (rev 2) ==

1 0x008 4 306 Lifetime Power-On Resets

1 0x010 4 1593 Power-on Hours

1 0x018 6 3211855176 Logical Sectors Written

1 0x020 6 117248637 Number of Write Commands

1 0x028 6 1302874550 Logical Sectors Read

1 0x030 6 31953300 Number of Read Commands

4 ===== = = == General Errors Statistics (rev 1) ==

4 0x008 4 0 Number of Reported Uncorrectable Errors

4 0x010 4 0 Resets Between Cmd Acceptance and Completion

6 ===== = = == Transport Statistics (rev 1) ==

6 0x008 4 2093 Number of Hardware Resets

6 0x010 4 3872 Number of ASR Events

6 0x018 4 0 Number of Interface CRC Errors

7 ===== = = == Solid State Device Statistics (rev 1) ==

7 0x008 1 0 Percentage Used Endurance Indicator

I hope it helps.

idata
Community Manager
155 Views

Well, 3211855176 written 512 byte sectors are 1.53 TiB. (With the assumption that a sector in fact has 512 bytes)

idata
Community Manager
155 Views

arcks, Did you try using the Intel SSD Toolbox, and check the SMART data with that?

idata
Community Manager
155 Views

Hello parsec,

That fancy Intel thing gives me the following SMART data:

Says exactly the same.

CFran6
Novice
155 Views

The ~2.4 times larger number of writes may be the result of some application caching file data in memory and doing frequent flushes (DBMS?). Some tool of choice might expose which processes produce these I/O rates. My favorite: http://processhacker.sourceforge.net/ http://processhacker.sourceforge.net/ - run as admin and observe the "Disk" tab.

You could also clear the Timed Workload attributes (see above) and observe attribute 227/E3 which reports the percentage of reads.

idata
Community Manager
155 Views

Hello Christian,

after executing 'smartctl -t vendor,0x40 /dev/sda' and waiting some minutes smartctl says:

3 Spin_Up_Time 0x0020 100 100 000 Old_age Offline - 0

4 Start_Stop_Count 0x0030 100 100 000 Old_age Offline - 0

5 Reallocated_Sector_Ct 0x0032 100 100 000 Old_age Always - 0

9 Power_On_Hours 0x0032 100 100 000 Old_age Always - 1602

12 Power_Cycle_Count 0x0032 100 100 000 Old_age Always - 308

170 Reserve_Block_Count 0x0033 100 100 010 Pre-fail Always - 0

171 Program_Fail_Count 0x0032 100 100 000 Old_age Always - 0

172 Erase_Fail_Count 0x0032 100 100 000 Old_age Always - 0

183 Runtime_Bad_Block 0x0030 100 100 000 Old_age Offline - 0

184 End-to-End_Error 0x0032 100 100 090 Old_age Always - 0

187 Reported_Uncorrect 0x0032 100 100 000 Old_age Always - 0

192 Unsafe_Shutdown_Count 0x0032 100 100 000 Old_age Always - 23

199 UDMA_CRC_Error_Count 0x0030 100 100 000 Old_age Offline - 0

225 Host_Writes_32MiB 0x0032 100 100 000 Old_age Always - 49181

226 Workld_Media_Wear_Indic 0x0032 100 100 000 Old_age Always - 65535

227 Workld_Host_Reads_Perc 0x0032 100 100 000 Old_age Always - 65535

228 Workload_Minutes 0x0032 100 100 000 Old_age Always - 65535

232 Available_Reservd_Space 0x0033 100 100 010 Pre-fail Always - 0

233 Media_Wearout_Indicator 0x0032 100 100 000 Old_age Always - 0

241 Host_Writes_32MiB 0x0032 100 100 000 Old_age Always - 49181

242 Host_Reads_32MiB 0x0032 100 100 000 Old_age Always - 20018

The values 226, 227 and 228 seem a little confusing.

Oh and apparently 6 GiB were written in the last two days. When this is really true, some process behaves really strangely.

I don't have a DBMS, this is just my desktop computer. IOtop says that chromium writes to disk frequently. I didn't think this was dangerous until now.

I deleted ~/.cache and mounted it as a 1 GiB tmpfs ramdisk, just in case it's really chromium's fault.

I'm a little worried about my SSD, this can't be healthy. :-(

CFran6
Novice
155 Views

The attributes 226, 227 and 228 are held at 65535 (0xffff) until the workload timer reaches some limit. The X25 G2 spec mentioned 60 minutes, the 320 spec does no longer document these attributes at all.

(BTW: The '-t vendor,0x40' option was added due to a request from an Intel engineer :-)

idata
Community Manager
155 Views

I tried it after 60 and 65 minutes, still 65535, but now it shows some values.

226 Workld_Media_Wear_Indic 0x0032 100 100 000 Old_age Always - 0

227 Workld_Host_Reads_Perc 0x0032 100 100 000 Old_age Always - 58

228 Workload_Minutes 0x0032 100 100 000 Old_age Always - 78

226 Workld_Media_Wear_Indic 0x0032 100 100 000 Old_age Always - 0

227 Workld_Host_Reads_Perc 0x0032 100 100 000 Old_age Always - 52

228 Workload_Minutes 0x0032 100 100 000 Old_age Always - 83

226 Workld_Media_Wear_Indic 0x0032 100 100 000 Old_age Always - 0

227 Workld_Host_Reads_Perc 0x0032 100 100 000 Old_age Always - 51

228 Workload_Minutes 0x0032 100 100 000 Old_age Always - 88

226 Workld_Media_Wear_Indic 0x0032 100 100 000 Old_age Always - 0

227 Workld_Host_Reads_Perc 0x0032 100 100 000 Old_age Always - 50

228 Workload_Minutes 0x0032 100 100 000 Old_age Always - 93

226 Workld_Media_Wear_Indic 0x0032 100 100 000 Old_age Always - 0

227 Workld_Host_Reads_Perc 0x0032 100 100 000 Old_age Always - 48

228 Workload_Minutes 0x0032 100 100 000 Old_age Always - 123

226 Workld_Media_Wear_Indic 0x0032 100 100 000 Old_age Always - 0

227 Workld_Host_Reads_Perc 0x0032 100 100 000 Old_age Always - 43

228 Workload_Minutes 0x0032 100 100 000 Old_age Always - 183

Steadily decreasing.

idata
Community Manager
155 Views

The internet browsers I've used will write to their cache all the time, you'll easily be able to see how much and how quickly using a ramdisk for the cache. It's also not unusual for a browser to have multiple files or caches they write to, in different locations. For example, I can move my Firefox cache to a different drive, and still have it write to the system drive. But all together the caches certainly do not add up to the amount of data that you are apparently seeing.

Assuming the hours on count is correct, that SSD has been on for a bit under 67 days total. At 1.5Tb written, that's almost 22.4GB a day. But don't feel to bad, given the Host Writes and Power on Hours Count of an Intel 510 SSD I have, I've written a bit over 70GB a day. I know how I've used that SSD, and I can't see how I've done that. IMO, some or all of this data must be wrong, or we don't know how it is accumulated/calculated by the SSD.

idata
Community Manager
155 Views

Hey parsec,

Thanks for these calmative words, I already questioned my sanity since outsourcing the chromium cache didn't show much effect. I guess I don't have to worry until the Media Wearout Indicator rises.

Reply