Internal HDD Problem

Black Hole said:
If the drive's controller detects a problem with a disk access, it swaps that area of the disk out and replaces it with one from a set of spares (built-in redundancy). When the spares run out, your data is at risk. The main reason for this reallocation is to overcome manufacturing defects which would otherwise trash the majority of the manufacturers output, but it also provides a useful self-healing mechanism for the warranted lifetime of the drive.

Random faults will occur from time to time. There should be no pattern to them, and a drive that reports occasional reallocations is not a worry. It is time to worry when large batches of reallocations are occurring - the drive surface is degrading, and data is on borrowed time.
Click to expand...
Ok. This is the mechanism I am aware of and what I was referring to by "survivable errors".

Black Hole said:
LBA 0 is a critical area of the disk which cannot be reallocated (or if it can, then only in a limited way). It contains the metadata necessary for low-level disk access, so if that is damaged the whole disk will become unaccessible.
Click to expand...
What does fixdisk do to repair LBA 0?
I did try to find info about fixdisk - is it the same as fsck on linux?
 
Trev said:
And after all, it's only tele.
Click to expand...
Try telling that to an eight year old who has just lost twenty episodes of Charlie and Lola. :eek:

Trev said:
Not like what you are likely to have on your PC (backed up I assume?)
Click to expand...
Yes agreed Trev, I am very careful about backing up my PC, to the extent of even keeping a separate backup usb drive at a relative's house in case my own house burns down!
 
Steve7777777 said:
What does fixdisk do to repair LBA 0?
Click to expand...
fixdisk doesn't do anything to correct disk faults, other than attempt to mark suspect sectors as bad and let the disk's built-in redundancy deal with it. fixdisk is primarily a file system analysis and correction tool, utilising fsck as you say (but made accessible for the rest of us by tuning fsck to the specific situation, instead of having to enter parameters ourselves).

A file system is essentially a structured database, and contains many different indexes to the data so that various modes of access can read or write data as quickly as possible. When one of the index mechanisms gets broken (corrupted), examination of the remaining indexing often provides enough redundancy to reconstruct the missing index.
 
Last edited:
Steve7777777 said:
but no-one has explained precisely what any of these numbers mean and why it is that you and Mike say scrap the disk and two other folk say don't give up.
Click to expand...
There is no precise explanation. Each drive failure will be somewhat different. I have over thirty years experience of using hard drives in personal computers (I currently own about 20 PCs) so hope that I have a sense of when it is worth trying to fix a drive (and many people give up too soon) and when it is better to give up and buy a new drive. In your specific case the reallocated sector count plus the problems with sector 0 lead me to the opinion that the drive is near end of life and whilst it may be possible to get it working again for a short while it will inevitably require replacing before long. I went through a similar process with a failing 1TB drive in our HDR-FOX T2 and as a matter of interest kept it running longer than most people would have but when it had reached the same state as yours the final deterioration was quite rapid. Your case is worse than ours as the area of the disk that seems to have a problem is in an area that is critical to the file system. My advice is to bite the bullet and replace the drive.
 
But there is no point asking for advice and then not taking it.

Would a small replaceable hdd for recording, backing up recordings to a large ssd, not be a better design for these recorders? The box could be designed to do the backup when a recording is finished, avoiding large numbers of writes to the ssd, and the ssd would also be protected from the 2 hour buffer and all the wear that entails.
 
Mike0001 said:
Would a small replaceable hdd for recording, backing up recordings to a large ssd, not be a better design for these recorders? The box could be designed to do the backup when a recording is finished, avoiding large numbers of writes to the ssd, and the ssd would also be protected from the 2 hour buffer and all the wear that entails.
Click to expand...
Replace the HDD with a large RAM and I would say yes, but these things are required to hit a price point on the high street rather than be technically perfect.
 
Well, it isn't relevant now, just the way things might be in the future, and prices are still falling.
 
What is really needed is some way of registration, so that, once machines have been registered to a user, they can transfer recordings between them, irrespective of make, and also whether Freesat or Freeview. Am I wishing for too much convenience here?
 
Well, mine decrypts automatically. I don't have to spend hours doing anything or even watching it. So it takes none of my time at all, let alone any needless time. What is needless time anyway. Hummy Arms, here we come.
 
I think Mike is talking in general, about what manufacturers ought to do (assuming they don't want to remove DRM) - but I think that's pie in the sky. And now they've seen what we can do (the royal we, obviously), they're going to make damn sure not to leave back doors open any more. Which is why I started the discussion about a DIY open standards PVR in the 'Arms.
 
Sorry I keep getting called away to other things, unfortunately fixing my PVR is low on the priorities atm.

Anyway thanks for all the additional info everyone. So taking all into account I am now looking to buy a new drive. :rolleyes:

Having searched online it seems the most recommended for hunax are seagate.

This page:
https://wiki.hummy.tv/wiki/2TB_Disk_Installation_Blog
has led me towards the Seagate ST1000VM002 Pipeline (I want to stick with 1TB):
https://www.amazon.co.uk/dp/B006WMB45K/
though I would buy it directly from amazon at £49.40 rather than risk a marketplace seller (previous bad experiences).

Any thoughts on the Seagate Pipeline? Is that still a good choice? (that wikki page gives some good arguments in its favour but was written more than a year ago)

Cheers
 
Steve7777777 said:
has led me towards the Seagate ST1000VM002 Pipeline (I want to stick with 1TB):
https://www.amazon.co.uk/dp/B006WMB45K/
though I would buy it directly from amazon at £49.40 rather than risk a marketplace seller (previous bad experiences).

Any thoughts on the Seagate Pipeline?
Click to expand...
We have the 2TB version (fitted when the Humax 1TB drive died after about 15,000 hours). It is now approaching 20,000 power on hours and so far there are no reallocated sectors and it has given no problems of any sort.
 
You must log in or register to reply here.
Back
Top