[BUG] Some drive types do not display value for Power on Hours about scrutiny

[AnalogJ]

Just looking at the data from the smartctl logs, it looks like nvme drives do not provide power on hours. Here's an example from an SCSI drive

SCSI output

time="2020-09-25T17:41:30Z" level=info msg="Executing command: smartctl -a -j /dev/sdk" type=metrics
{
  "json_format_version": [
    1,
    0
  ],
  "smartctl": {
    "version": [
      7,
      0
    ],
    "svn_revision": "4883",
    "platform_info": "x86_64-linux-4.19.107-Unraid",
    "build_info": "(local build)",
    "argv": [
      "smartctl",
      "-a",
      "-j",
      "/dev/sdk"
    ],
    "exit_status": 0
  },
  "device": {
    "name": "/dev/sdk",
    "info_name": "/dev/sdk",
    "type": "scsi",
    "protocol": "SCSI"
  },
  "vendor": "SEAGATE",
  "product": "ST9300503SS",
  "model_name": "SEAGATE ST9300503SS",
  "revision": "MS01",
  "scsi_version": "SPC-3",
  "user_capacity": {
    "blocks": 585937500,
    "bytes": 300000000000
  },
  "logical_block_size": 512,
  "rotation_rate": 10000,
  "form_factor": {
    "scsi_value": 3,
    "name": "2.5 inches"
  },
  "serial_number": "6SE0N87PXXXXXXX",
  "device_type": {
    "scsi_value": 0,
    "name": "disk"
  },
  "local_time": {
    "time_t": 1601055690,
    "asctime": "Fri Sep 25 17:41:30 2020 America"
  },
  "smart_status": {
    "passed": true
  },
  "temperature": {
    "current": 47,
    "drive_trip": 68
  },
  "scsi_grown_defect_list": 0,
  "power_on_time": {
    "hours": 168,
    "minutes": 42
  },
  "scsi_error_counter_log": {
    "read": {
      "errors_corrected_by_eccfast": 748616,
      "errors_corrected_by_eccdelayed": 0,
      "errors_corrected_by_rereads_rewrites": 0,
      "total_errors_corrected": 748616,
      "correction_algorithm_invocations": 748616,
      "gigabytes_processed": "505.278",
      "total_uncorrected_errors": 0
    },
    "write": {
      "errors_corrected_by_eccfast": 0,
      "errors_corrected_by_eccdelayed": 0,
      "errors_corrected_by_rereads_rewrites": 0,
      "total_errors_corrected": 0,
      "correction_algorithm_invocations": 0,
      "gigabytes_processed": "2763.970",
      "total_uncorrected_errors": 0
    }
  }
}

[AnalogJ]

We should investigate the results using the -x flag, which provides additional device data. #9

@teambvd can you try running the following command and paste the results below?
docker exec scrutiny smartctl -x -j /dev/sdi

Thanks!

[teambvd]

That drive is a SAS drive (not NVMe) - the above command does give the power on hours at the very end:
"power_on_time": {
"hours": 39403,
"minutes": 32

NVMe drives display power on hours properly as is:

Full output of the docker exec command:
docker exec scrutiny smartctl -x -j /dev/sdi

{
  "json_format_version": [
    1,
    0
  ],
  "smartctl": {
    "version": [
      7,
      0
    ],
    "svn_revision": "4883",
    "platform_info": "x86_64-linux-4.19.107-Unraid",
    "build_info": "(local build)",
    "argv": [
      "smartctl",
      "-x",
      "-j",
      "/dev/sdi"
    ],
    "exit_status": 0
  },
  "device": {
    "name": "/dev/sdi",
    "info_name": "/dev/sdi",
    "type": "scsi",
    "protocol": "SCSI"
  },
  "vendor": "TOSHIBA",
  "product": "MG03SCA300",
  "model_name": "TOSHIBA MG03SCA300",
  "revision": "DG04",
  "scsi_version": "SPC-4",
  "user_capacity": {
    "blocks": 5860533168,
    "bytes": 3000592982016
  },
  "logical_block_size": 512,
  "rotation_rate": 7200,
  "form_factor": {
    "scsi_value": 2,
    "name": "3.5 inches"
  },
  "serial_number": "Z4K0A12YFVL9",
  "device_type": {
    "scsi_value": 0,
    "name": "disk"
  },
  "local_time": {
    "time_t": 1601151939,
    "asctime": "Sat Sep 26 20:25:39 2020 America"
  },
  "smart_status": {
    "passed": true
  },
  "temperature": {
    "current": 28,
    "drive_trip": 65
  },
  "scsi_grown_defect_list": 0,
  "scsi_error_counter_log": {
    "read": {
      "errors_corrected_by_eccfast": 0,
      "errors_corrected_by_eccdelayed": 9,
      "errors_corrected_by_rereads_rewrites": 9,
      "total_errors_corrected": 9,
      "correction_algorithm_invocations": 15,
      "gigabytes_processed": "5674097.674",
      "total_uncorrected_errors": 0
    },
    "write": {
      "errors_corrected_by_eccfast": 0,
      "errors_corrected_by_eccdelayed": 0,
      "errors_corrected_by_rereads_rewrites": 0,
      "total_errors_corrected": 0,
      "correction_algorithm_invocations": 0,
      "gigabytes_processed": "102102.605",
      "total_uncorrected_errors": 0
    },
    "verify": {
      "errors_corrected_by_eccfast": 0,
      "errors_corrected_by_eccdelayed": 1,
      "errors_corrected_by_rereads_rewrites": 1,
      "total_errors_corrected": 1,
      "correction_algorithm_invocations": 1,
      "gigabytes_processed": "669629.192",
      "total_uncorrected_errors": 0
    }
  },
  "power_on_time": {
    "hours": 39403,
    "minutes": 32
  }
}

[AnalogJ]

Oh apologies. I should have taken a closer look at the drive type.
That's fantastic though, I'm glad the -x flag works. Judging by its description, it should return both the device data and the smart data for all drive types, but I've been bitten by weird inconsistencies in the past.

Let me add the flag the collector in a new branch and spin up a new docker image for you to try out.

[AnalogJ]

Alright, I got the image built.

can you run the following image: analogj/scrutiny:x_flag

docker run -it --rm -p 8080:8080 \
-v /run/udev:/run/udev:ro \
--cap-add SYS_RAWIO \
--device=/dev/sda \
--device=/dev/sdb \
--name scrutiny \
analogj/scrutiny:x_flag

[mglubber]

Not the original poster, but I had a similar issue with an Intel 520 SSD - for some reason smartctl -a reports an absurd value for the drive. It's bit older, but I don't think it's been powered on for 105 years. Using the 'x_flag' docker images returns a much saner value of 3 years.

[teambvd]

@AnalogJ I think that cleared up the last of my issues! To summarize the configuration, in case it helps with your documentation:

• Poweredge r720xd
  • 2x 2.5" 10k SAS FDE drives on secondary internal midplane (separate bus, same chipset)
  • 10x 3.5" SAS and 2x 3.5" SATA on primary internal backplane
  • 1x Intel AIC NVMe (PCIe add-in-card)
• 12x external 3.5" SAS FDE drives connected via LSI 9207-8e in an EMC KTN-STL3 disk shelf

Basically, this is about the biggest mix and match one could expect to have, at least as far as protocols are concerned. The only way I could see this getting confused now is if there were some host multipathing issues, but that's outside the scope of this product (and wouldn't be your problem anyway).

Great friggin work! Do you need the logs for anything, or...? I didn't figure you would if it all worked as designed, but perhaps you've other needs for em I don't know about...?

[AnalogJ]

Hey @teambvd
That's great to hear. Thanks for all the additional information about your setup.
I don't need any log files actually, it sounds like everything is working as intended :)