Sometimes, in order to find out that the HDD is dying, you need to react to errors detected by the application software

I made it a rule to always install all OS updates, possibly with a slight delay (up to a week), I also installed a special application (APP Center) from the motherboard manufacturer, which monitors the update of drivers and some programs, and also includes such useful programs as , for example, setting up the fans or updating the UEFI (in 2018, I had to update 3 times due to identified problems in Intel processors). In this application, until recently, everything worked as expected. Every day I checked to see if there was anything new and usually immediately set it. That was until recently. When the sound card software updates arrived (it completely uninstalled the classic application and installed a new one), I waited a bit and decided to install. The APP Center’s small problem is that it offers to install or update a number of programs that I don’t need, such as an antivirus, and the purchased one is quite suitable for me, so before upgrading I always go over the list and remove unnecessary daws. So it was this time. But something went wrong.



In general, there are usually two tabs, on the bottom there is an “Update” button, on the other “Install”, on each “Cancel”. Only the contents of one tab are visible at a time. That day a third appeared. It would be possible now to go back to the moment before the incident and post a screenshot here, and then go back to the current moment (I have not yet deleted the backups, although of the old format), but I try not to use Acronis True Image recovery without the need to - even try to first problem solve in the usual ways. And even more so, I do not jump into various states of the system because of the desire to take a screenshot (although recovery occurs quickly and starts directly from a working OS, you only need to allow a reboot). Therefore, I rely on memory. And she can let you down: that day I ran through the tabs several times. He did, it seems, everything was logical, but something did not take into account.



Before updating, I did not look at the third tab, where the application also put daws opposite the driver updates to the network card and HDD. If I wanted to update them, then not on this day. Therefore, when I saw something unexpected, I immediately stopped updating. Well, since I saw that some of the drivers were installed, and some were not, I rolled back to the morning copy of the system using Acronis True Image 2019, which I sometimes did in such cases in order to update tomorrow, but in the meantime do other things. I worked all day without problems, and the next morning, Acronis did not create a backup copy due to hard disk errors. The obvious solution is to run chkdsk. Verification failed. Then chkdsk c: / scan / v / r. There were errors in the Chrome cache, in order to fix it, it was necessary to reboot. I agreed. This was the last thing that could be done.



At boot, the system fell into the BSOD and began to reboot to the next BSOD and so on in the loop. The desktop did not reach. Not a problem: my backups are stored on an external USB HDD on which the Acronis Survival Kit is installed, which at boot allows you to start the recovery of the backup in WinRE. Failure. The program recommended booting from a Linux Acronis flash drive and continuing recovery. But I didn’t do this flash drive in advance. But there was a RecoveryDrive flash drive, it booted from it (in order to use someone else's computer, I would have to wait - it was early morning) and put a clean system on the computer. Installed True Image and created a bootable USB flash drive with Linux version of this software.



I restored a copy of the system the night before the incident and again checked the hard drive - the same errors. I rummaged in the System magazine, found that the problem was most likely in one file system filter (errors occurred immediately after loading it, and the date was already too old). This time, the APP Center (in this copy) no longer offered to upgrade, so it recovered to the last copy before the incident. Successfully updated all the drivers, checked everything again. Additionally installed smartmontools. It turned out that during this time SMART managed to reassign more than 2200 sectors (almost 100%), in general, the HDD is corrupted, although the problem is not in defects on the disk surface, but only in software. But at least it worked. Until I bought a subscription to Acronis True Image 2020.



Updated and again could not make a copy. More precisely, sometimes a copy was made, sometimes not due to HDD errors. Wrote Acronis technical support (regular tools did not reveal problems) and continued trying to find what does not work this time. I also checked hidden drives with chkdsk - no problem. I got to SeaTools - software from the manufacturer of the HDD just for such cases. So it revealed the problem - the DST test did not pass. It was proposed to create a Linux flash drive and continue the repair by booting from it. Which I did. DST also did not go through to the end, launched a full test, and then a short repair. After that, the DST passed. And Acronis True Image 2020 began to make copies without problems. But questions remained.



After the repair, smartctl showed that the number of reassigned sectors did not increase, chkdsk (including hidden disks) also did not show an increase in the number of bad sectors, it seems that the problem has not been completely resolved and may appear in the future. And this means that you probably have to change the HDD. By the way, Acronis support answered on Monday (I sent the request on Saturday), but I had already solved the problem (although it’s already impossible to buy Acronis True Image for rubles in Russia, but there is still support for individuals, I’ve got used to it since 2010 I’m not going to change software, I’ll pay in Euros now).



Perhaps data recovery specialists will be able to clarify whether my actions and conclusions were correct. I must say right away that now in Windows 10 the restoration of a knownly working VSS copy does not already give confidence in its further normal operation.



UPD: Sergey_datex writes that the disk is dead. Probably it is.

UPD: Corrected the title after oller comment

UPD: As some conclusion to the article, it seems to me that this comment of mine may come up.



All Articles