Drive reaches 100% and PC freezes
Troubleshooting with hardware (SSD)
The drive is operating at 100% and there are almost no reads or writes.
When this happens, your PC may freeze or become unresponsive.
Especially when the target is C drive, there are cases where you cannot even shut down the OS.
This phenomenon is often caused by loading to a bad cluster, which occurs mainly on SSDs.
SSDs containing NVMe get stuck at 100% utilization and no IO when the controller attempts to read a bad cluster (physically corrupted storage area) on request.
This may be resolved by timeout if you leave it for a while if you do not access the area again, but if the file required by the OS is in the bad cluster, control may not be returned. There are many.
RAID controllers that include a cache, such as the AMD-RAID, take a particularly long time to time out, and the IO will remain in a completely zero state.
The developer who is writing this has been experiencing this problem since he created the information acquisition program for AMD-RAID, and suspecting that the program he created was the cause, he changed the conditions on two PCs to isolate the cause. was doing.
As a result, I found that this phenomenon still occurs even if I completely eliminate the use of AMD_RC2t7 .
In the end, I were able to confirm that 「when you try to read a specific part of a specific file, it is reproduced 100%」, and this problem was caused by a bad cluster on the SSD (access to a storage area with physical problems). It turned out that this was the cause of the problem.
Added below 2022/10/27:
After this, I completely removed the problematic drive and changed from SSD to HDD, and started testing again by restarting the use of AMD_RC2t7, and even though there was no problem with the drive after testing, similar The problem has been confirmed to be reproduced.
Apparently, this is a different issue than the drive problem, and it seems that AMD_RC2t7 is also the cause of unstable operation in some environments.
It has been more than 5 weeks since I started operational testing of 0b6 Rev.1 (thilmera7 0b176) of AMD_RC2t7, which is an experimental fix after I became aware of this problem, but this problem has never occurred on two PCs. It did not occur.
For the AMD RAID function of AMD_RC2t7 and thilmera, please use the corresponding version AMD_RC2t7 0b6 Rev.1 or later.
thilmera 7 is compatible with 0b176 or later.
Versions of AMD_RC2t7 known to have issues・AMD_RC2t7 : 0b1 to 0b6
・ thilmera 7 : 0b171 to 0b175 Rev.2
Version AMD_RC2t7 resolved the issue・AMD_RC2t7 : 0b6 Rev.1～
・ thilmera 7 : 0b176 ~