Separate names with a comma.
Discussion in 'DIRECTV HD DVR/Receiver Discussion' started by hoyty76, Dec 7, 2008.
How do you run that test?
Also, does anyone know where I can pre-order this drive?
Not sure about this one, but most drives I've purchased came with a utilities disk that contained low-level diagnostics. /steve
From reading this thread images of Tim Allen and a monster DVR with fans and HDDs galore come to mind .
I've always liked a software program called Spinrite for lowlevel HD diags.
I've used SpinRite since I bought my first HR10-250 on April 12, 2004 and it works great. I completely forgot all about it but I used it to remove bad sectors after I started having problems with pixellation and macro blocking and slowing down and sluggish remote commands.
That's a pretty high failure rate. As a computer consultant that works on hundreds of systems per year, I stopped doing the "long" formats a few years ago. I can name a number of things I'd like to do rather than wait for a drive to format. I do make sure there is a quality backup system in place though.
If I'm concerned about it I'll do a chkdsk/fsck with sector checking after the system install. Modern drives do have spare sectors that are used to remap bad sectors dynamically. But it's no guarantee that data won't be stored in a sector before it fails and gets remapped(and potentially corrupted). I certainly wouldn't worry about it in a DVR application.
Block size is another factor in how fast that format goes, with smaller block sizes having more overhead and taking longer. I wonder if the OS in the HR20 accounts for overall drive size and adjusts the block size accordingly. If it doesn't that might account for the reports I've seen where the system slows down when people install larger drives.
The drivers know what to do with some brands' hard drive intelligence including remapping. I presume it also can do some bad sector flagging on less familiar drives.
There is a built in test that can be run to verify the file system that is not destructive. Tho it can take some time on a 750gb drive. Don't even want to know how long it can take on a 2TB drive...
(Ok, actually I do want to know. I'd be very happy to test a drive if someone gave me one.)
Wll, I should have 3 WD20EADS Drives in my possession shortly but I think I can test them thoroughly by myself but thanks TOM for the thought and help in testing these puppies. I have been waiting a year for these to come out and I can hardly wait.
The sector remapping/sector sparing is handled by the drive firmware and is transparent to the OS/drivers.
Yes, that's not to say that the filesystem can't implement it's own methods when controller errors are reported.
The old Novell Netware OS had a separate Hotfix partition where it would remap blocks when there were read errors on a block. NTFS has cluster remapping. It's just happening on a different level.
It would make sense that the firmware had a major role, as that allows the manufacturer to address changes, problems, etc. that way...
Tom remember this?
Number of cylinders: 604
Number of heads: 4
Number of sectors per track: 17
Number of bytes per sector: 512
Do you want to enter any bad sectors y/n: n
Lowlevel format drive? (warning this will erase the drive) y/n: y
sorry, too far off topic?
Spinrite can also do sector remapping, which is one of it’s selling points, especially verses NTFS. Big difference between 512 bytes and 64 Kbytes for a cluster back in the NT 4.0 days. These days with lba48 we are back down to 4 Kbytes but 512 bytes is still 1/8 the loss.
I can confirm it is non-destructive. I've run it overnight and my recordings were fine the next day. During the boot process, press SELECT when you see the "Running receiver self-check" message on-screen. You'll enter the diagnostic menu system. /steve
Yes I remember...
One of the first harddrives I ever bought was supposed to be a 30meg. The tech entered the parameters for a 30, but it was really a 20. All was fine until I reached the 20 meg point. Computer store was out of business by that time.
This is a bit of a sidebar, but I had D* replace a hard drive in a Tivo. The old hard drive held 30 hours. The one they replaced it with only had 20. I sent it back and the drive said it had 30. Then when I got to 20 hours it crashed and came back up that it only had 20 hour capacity. D* entered parameter to fool the Tivo. It took 3 tries, but they finally gave me the correct drive. Still running 7 years later.
When my 1TB drive dies, I'll get one of the 2TB ones.
Isn't 2TB the largest that the HR* series recognizes?
Not at all. Calling the LLF ROM routine from debug is still relevant today.
Why just the other day I had to slam my hard drive on the desk because the heads became magnetized and were stuck to the platters. The darn thing wouldn't spin up
HR20 currently limited to 2TB, HR21/22/23 limited to 16TB.
(and I remember those commands too. As well as formatting dectapes via toggle switches...)
Cool. I hope my 1TB lasts until the 8 or 16TB drives come out