Possible incorrect Write/Read numbers

[konn-neko]

when checking via crystal disk info for this drive https://www.silicon-power.com/web/product-ace_a55

(refrence image using same type of drive)

ther tbw is around 125

QDiskInfo_sde.json

[konn-neko]

posibly related to #50

[edisionnano]

Perhaps it is wrong perhaps it isn't. Your power on hours are equal to roughly 2 years of consecutive use. I see that you have a lot of drives and I'm not familiar with you usage patterns but for such a small capacity drive and this many hours of usage less than 2 TB of reads is not a very believable number. As for the rated (key word here) TBW, the drive won't die if you exceed them but your warranty is probably void.

[konn-neko]

Perhaps it is wrong perhaps it isn't. Your power on hours are equal to roughly 2 years of consecutive use. I see that you have a lot of drives and I'm not familiar with you usage patterns but for such a small capacity drive and this many hours of usage less than 2 TB of reads is not a very believable number. As for the rated (key word here) TBW, the drive won't die if you exceed them but your warranty is probably void.

this is the same exact one crystal disk info shows in my pc

hence why i think its off by a bit, on my maion pc i dont use the drive much hence the low number

[windwaker87]

I bought a new 8TB drive, packed out of the box and copied my existing old drive to the new one.

I suspect this app is showing gigabytes as terabytes. Once it reached 1 terabyte of data written, it started showing correctly, but while it was less than 1TB of data written, it showed hundreds of TB. Same with read data value. It's currently showing 945TB but that is impossible for a new drive. Total Host Reads value was steadily increasing by hundreds of TB while data was being synchronized with rsync.

According to smartctl:

=== START OF INFORMATION SECTION ===
Device Model:     ST8000DX001-3GK101
Firmware Version: FC04
User Capacity:    8.001.563.222.016 bytes [8,00 TB]
Sector Sizes:     512 bytes logical, 4096 bytes physical
Rotation Rate:    7200 rpm
Form Factor:      3.5 inches
Device is:        Not in smartctl database 7.3/5528
ATA Version is:   ACS-4 (minor revision not indicated)
SATA Version is:  SATA 3.3, 6.0 Gb/s (current: 6.0 Gb/s)
SMART support is: Available - device has SMART capability.
SMART support is: Enabled

Device Statistics (GP Log 0x04)
Page  Offset Size        Value Flags Description
0x01  =====  =               =  ===  == General Statistics (rev 1) ==
0x01  0x008  4               2  ---  Lifetime Power-On Resets
0x01  0x010  4              16  ---  Power-on Hours
0x01  0x018  6     11927823896  ---  Logical Sectors Written
0x01  0x020  6        12658144  ---  Number of Write Commands
0x01  0x028  6         1352660  ---  Logical Sectors Read
0x01  0x030  6           86689  ---  Number of Read Commands
0x01  0x038  6               -  ---  Date and Time TimeStamp

512 logical sector size and 1352660 logical sectors read - it should show 660,48MiB of Total Host Reads (if I'm calculating correctly). If I calculate manually Total Host Writes I get 5.55TiB so it appears to be showing correct data for writes since it already is way past 1TB of data written.

[windwaker87]

I found the suspect code: in src/mainwindow.cpp starting with line 609

                } else if (attrObj["name"] == "Total_LBAs_Written") {
                    int logicalBlockSize = localObj["logical_block_size"].toInt();
                    qlonglong lbaWritten = attrObj["raw"].toObject()["value"].toVariant().toLongLong();
                    qlonglong gigabytes = (lbaWritten * logicalBlockSize) / 1e9;
                    int gigabytesInt = static_cast<int>(gigabytes);
                    if (!gigabytesInt) {
                        gigabytesInt = static_cast<int>(lbaWritten);
                    }
                    totalWritesInt = gigabytesInt;
                }

If after calculations the gigabytes value is less than 1 (zero in my case) it defaults to lbaWritten value thus inflating the value a billion times. And I would suspect it to be the same for Total_LBAs_Read.

Allowing 0 gigabytes value to exist, or replacing the logic to use megabytes instead of gigabytes would fix that.

[windwaker87]

edit: after looking more closely to the screenshots posted, @konn-neko SSD was connected over SATA, so it is more likely that his drive would be affected by the code mentioned in my previous comment.

Also very interesting code in src/mainwindow.cpp starting with line 679.

We see dataUnitsWritten and dataUnitsRead being converted to gigabytes, but actual code shows it's being converted to megabytes and reported as gigabytes. This code in particular works for NVME SSD's so it explains previous comments from @konn-neko about SSD values being incorrectly reported. It's possible that the dataUnitSize is different for each SSD. I can't find it in the output JSON or using smartctl. But it appears to be reported in 512KB for my own SSD, even though information on the web suggests it should use the logical_block_size value which is 512B.

It's also strange to see a hardcoded 512 being used here instead of getting logical_block_size from the device because they might be using 4/8/16 KB blocks.

    } else {
        for (auto smartItem = nvmeLog.begin(); smartItem != nvmeLog.end(); ++smartItem) {
            QString key = smartItem.key();
            QJsonValue value = smartItem.value();
            if (key == "data_units_written") {
                double dataUnitsWritten = value.toDouble();
                double gigabytes = (dataUnitsWritten * 512) / 1'000'000;
                totalWritesInt = static_cast<int>(gigabytes);
            } else if (key == "data_units_read") {
                double dataUnitsRead = value.toDouble();
                double gigabytes = (dataUnitsRead * 512) / 1'000'000;
                totalReadsInt = static_cast<int>(gigabytes);
            } else if (key == "percentage_used") {
                int percentageUsed = 100 - value.toInt();
                percentage = QString::number(percentageUsed) + " %";
            }
        }
    }

[windwaker87]

After making the changes to convert into megabyte logic + fixing the calculations I get the correct results for all my drives:

[edisionnano]

After making the changes to convert into megabyte logic + fixing the calculations I get the correct results for all my drives:

Nice catch, I messed it up at some point probably, open a pull request

[windwaker87]

Nice catch, I messed it up at some point probably, open a pull request

Sorry, I don't have much experience with github. Most of my work was in gitlab. I don't have permission to push my new branch to this project and so can't make a pull request. Unless I'm doing something wrong...

[edisionnano]

Nice catch, I messed it up at some point probably, open a pull request

Sorry, I don't have much experience with github. Most of my work was in gitlab. I don't have permission to push my new branch to this project and so can't make a pull request. Unless I'm doing something wrong...

The way it works on GitHub, is you click the fork button which creates a fork of the project on your account. You then make changes to your fork and GitHub will prompt you to create a new pull request.

[windwaker87]

The way it works on GitHub, is you click the fork button which creates a fork of the project on your account. You then make changes to your fork and GitHub will prompt you to create a new pull request.

Thank you for the explanation. I've submitted a pull request #66

[konn-neko]

@windwaker87 thank you for explaining and digging into this, life caught up and i forgot i made this, honestly love you took the time for this <3

[edisionnano]

Since I don't have windows, or many drives. I want as many people as possible to try #66 and compare with Crystal. Thank you