bill3579 - 26 minutes ago »
However, the HDD test in Settings - Data Storage shows a disk failure, error 81, and a suggestion that format should be used to recover the data storage. The raw values in the drive diagnostics are 16 (buff), 9560, 27, 2 (orange) and 2 (orange).
I will run Fix Disk again
OK you have certainly improved things but as you suspect things are still not quite right. See how the next run of fix-disk goes.
You are correct, I am stuck in a loop and cannot see how to escape from it.
I had a look at the link and can see the outline of what I have to do, but not the details.
The loop I am stuck in is
Checking to see if this block is in use...
debugfs 1.42.10 (18-May-2014)
Block 225443839 is not in use
Dev: /dev/sda LBA: 1805655232
LBA: 1805655232 is on partition /dev/sda2, start: 2104515, bad sector offset: 1803550717
dumpe2fs 1.42.10 (18-May-2014)
Using superblock 0
Block size: 4096
LBA 1805655232 maps to file system block 225443839 on /dev/sda2
I'd really welcome more advice.
Oh dear. What is the model of the hard drive (it is shown in the second line of the Hard Drive diagnostic produced by the Humax) and what is the Physical sector size (seventh line down)?
Device Model ST1000VM002-1CT162
Sector Sizes 512 bytes logical, 4096 bytes physical
From post #62 it is an AF disk. It looks like the problems are with two physical sectors (equivalent to 16 logical sectors) at LBA 1805655232 and 1805655240.
In my post (#60 in this thread) there is a link to a thread at Hummy.TV. It is post #6 in that thread that has the instructions.
You need to do the following from Telnet:
Unmount the partition in question. In your case:
umount /dev/sda2
Then try reading the first logical sector in question:
hdparm --read-sector 1805655232 /dev/sda
Fix-disk has probably fixed this first sector so you will likely get lots of zeroes returned. If it is OK then try to read the next sector:
hdparm --read-sector 1805655233 /dev/sda
If you get a read error you need to overwrite this sector:
hdparm --write-sector 1805655233 --yes-i-know-what-i-am-doing /dev/sda
Then read the next sector and overwrite if you get a read error. Doing this up to LBA 1805655239 should fix the first physical sector. As you also seem to have a bad sector starting at LBA 1805655240 you will need to follow the same procedure from there up to LBA 1805655247, using the command to write the sector if you get a read failure. When you stop getting read errors (probably at LBA 1805655248) you can stop. Then reboot and see if you can do a disk check: start with the quick Humax one and see if it completes without giving you an error code. If OK it would be a good idea to run fix-disk again.
Thanks, I am going to try it, fingers crossed
I fell at the first hurdle with this result.
humax# umount /dev/sda2
umount: can't umount /mnt/hd2: Device or resource busy
I cannot see what HDR is doing to be busy, not playing or recording. Will resume in the morning.
Try doing it from maintenance mode. Some process is blocking the unmount. If the unit is connected to an aerial it is probably the unit writing to the timeshift buffer.
I worked from the Maintenance menu and read all sectors from 1805655232 to 1805655247 inclusive, one failed and was overwritten.
There were still disk errors so I ran Fix Disk and found myself still in the same loop, so I worked from the maintenance menu again to repeat reading the same sectors and to overwrite one that failed.
There were still errors; the Reallocated_Sector_Ct raw value was 16 so I ran fix disk again.
This time it completed without any reported problem but when I pointed Chrome at the HDR IP, a fault was reported, the Reallocated_Sector_Ct raw value is still 16 and the first error LBA is 1805665232 as before.
I hope someone familiar with linux syntax will create a module to execute the procedure for reading eight consecutive sectors, and overwriting any that return an error, starting from an input LBA. I taught myself to program in Visual Basic 6 about 20 years ago and, as I recall, it should not be very difficult.
So will the short disk check (the standard one that can be done using the remote control) run now without errors? You should have reallocated sectors but the smart statistics lines 187, 197 and 198 should be clear. If the short disk check still fails and you get a fault at a particular LBA value then you still have a bad sector that needs reallocation.
You must log in to post.
