TiVo Community Forum

TiVo Community Forum Archive 2
Covering threads with a last post date between
July 1, 2004 and December 31, 2005.
THIS IS A READ ONLY SITE
 


 

SEARCH  |  ARCHIVE 1 MAIN SITE

 
Forum Jump
 
Thread Tools Search this Thread Display Modes
Old 12-14-2005, 12:12 AM   #1 (Print)
smithcferg
Registered User
 
Join Date: Mar 2003
Posts: 13
Decompression error on restore

I am upgrading original 40 gig in s2 to 120 western digital drive.

I have made many backups and have been unable to successfully restore, except
I made eight backups, using compression levels -7 through to no compression. Almost all of the time the restores failed with "decompression error .x%" Once a compressed backup was successful, but it failed when put into the Tivo (wouldn't go past the grayscale "powering up" screen.) When I tried to re-restore that same backup, it failed with a decompression error.

The noncompressed backup came up with this error message upon completion:
mfs_read_inode: Inode 64930 corrupt, trying backup. I don't know yet if this is a fatal error.

This is the current mfstools2 boot cd that I am using, downloaded from Tyger's site.

The computer I am using is an Athlon 1.3 xp.

I came across a thread from this site that is now archived (6/2004) where Tyger was getting feedback from people who were complaining about getting constant decompression errors.

Does anyone know what is up, what to look for, workarounds, etc.

I am getting really tired of trying to make a working backup. I have spent all day on it!

Thanks to all for your help.

Smithcferg

P.S. I know the one thing I have not done is check the drives out to see if they are bad. That is next on my list.
smithcferg is offline Report Bad Post Report Post
Old 12-15-2005, 03:31 PM   #2 (Print)
smithcferg
Registered User
 
Join Date: Mar 2003
Posts: 13
Well, just wanted to let people know where this was goinog, and ask for input.

I have check the drives, they pass the five minute test from western digital diagnostics.

I also found out that when I try mfsrestoring an uncompressed backup multiple times that often I get different mfs_read_inode errors. Is this normal? I would expect the same errors each time.

I tried running the mfsbackup -Tao |mfsrestore -s 127 -xpzi command, which worked.

But when I tried the newly copied drive in the Tivo, again it failed to boot (though this time it did make it up to the "Almost there. Just a few more minutes" screen.
I double checked my original drive in the Tivo, it works fine. I have not tried using dd to avoid mfsbackup/restore commands altogether. First I will try a different upgrade hard drive, a 160 Gig I have lying around somewhere.

Anybody got any other helpful ideas?

SIncerely,
smithcferg
smithcferg is offline Report Bad Post Report Post
Old 12-15-2005, 05:36 PM   #3 (Print)
smithcferg
Registered User
 
Join Date: Mar 2003
Posts: 13
Ideas anybody?

Could really use an expert's help!

Thanks
smithcferg is offline Report Bad Post Report Post
Old 12-15-2005, 08:20 PM   #4 (Print)
HomeUser
Registered User
 
HomeUser's Avatar
 
Join Date: Jan 2003
Location: A2 Michigan
Posts: 529
Quote:
I tried running the mfsbackup -Tao |mfsrestore -s 127 -xpzi command, which worked.
Your command is missing a couple of '-'
if your drives are installed Original TiVo drive primary master and the new drive is secondary master
mfsbackup -Tao - /dev/hda | mfsrestore -s 127 -xpzi - /dev/hdc

for the backup compression problem I don't think 7 is valid. what are the other backup command options you are using? There could be memory or hard-ware problems with the PC.
HomeUser is offline Report Bad Post Report Post
Old 12-16-2005, 12:29 AM   #5 (Print)
smithcferg
Registered User
 
Join Date: Mar 2003
Posts: 13
Thanks for your input, HomeUser

I did use a command like yours for the piped backup/restore copy:
mfsbackup -Tao - /dev/hdb | mfsrestore -s 127 -xpzi - /dev/hde

I used standard commands for backups too:
mkdir /mnt/dos
mount /dev/hda2 (hda2 was a fat32 partition with plenty of room) /mnt/dos
mfsbackup -f9999 [-n] (depending on compression level I tried, and I tried none, -1, -2, -3, -4, -5, -6, -7) -so /mnt/dos/tivo.bak /dev/hdb
umount -f -a -r
ctl+alt+del

and for restore:

mkdir /mnt/dos
mount /dev/hda2

either
mfsrestore -xpzi /mnt/dos/tivo.bak /dev/hde

or
mfsrestore -pzi /mnt/dos/tivo.bak /dev/hde
mfsadd -x /dev/hde

or

mfsrestore -bpzi /mnt/dos/tivo.bak /dev/hde
mfsadd -x /dev/hde


or
mfsrestore -i /mnt/dos/tivo.bak /dev/hde
mfsadd -x /dev/hde

I also slowed memory speed way down just in case. And I wasn't using byte swapping, but did have DMA enabled.

The motherboard is a KK266R by IWIll, with 764 mb memory and Athlon xp 1.3 gighz cpu.

I didn't run into any problems actually executing commands as detailed in Hindsdale's how to. I have typed the dumb things over so many times that I have got them pretty well memorized.

It did seem problematic that I could restore from an uncompressed backup multiple times and get different mfs_read_inode errors each time.

I ran across this in a post:
An interesting note I received from someone:
"It turns out that when I went to my Parents house over the weekend to set up their Tivo and I was using one of his Pentium III machines to do the setup mfsrestore command failed to do a Decompression Error around 42%. I was a bit perplexed since I was using the same disks that I had used for the other two. I then went to a AMD 2.3 machine that he happened to have there and I got the same message. At this point I tried to run a diagnostic on the hard drive to see if it had a corrupt cluster and it came back in good condition.
I searched the newsgroups and found that someone believes there is a issue with some motherboards and the decompression routine. So finally we went to a Pentium Dell machine and it worked fine."

Also ran across this archived thread with input from Tyger that talks about a possible decompression problem, and I was wondering if it was ever resolved?
http://archive.tivocommunity.com/ti...c/170647-1.html However, this doesn't make sense as I am getting problems with straight non-compressed restores also.

Usually I have been running one hard drive off of a raid controller. The boot cd doesn't complain about it, and the sizes for the hard drives are correct, but I will try a piped backup/restore copy using just the regular ide. Who knows, maybe it will work. Hope springs eternal!

I had another hard drive around somewhere that I wanted to try instead of this 120 gig, but right now I can't find it.

Continued thanks for your input.

Sincerely,

Smithcferg
smithcferg is offline Report Bad Post Report Post
Old 12-16-2005, 04:43 PM   #6 (Print)
smithcferg
Registered User
 
Join Date: Mar 2003
Posts: 13
Thumbs up the smell of victory and odors from the battle

I did try the regular (non-raid) ide connections of the Iwill KK266R motherboard, and the copy completed successfully when trying a piped mfsbackup/restore copy, and as usual the Tivo didn't boot up right. (continual reboot on the power up screen)

So I took this same 120 gig hard drive to another computer with the soyo kt400 Dragon Platinum motherboard. Previously I had tried doing restores from this computer and they had failed using the onboard raid connector. This time I did a piped mfsbackup/restore copy (with same parameters as last attempt) using only the normal (non-raid) connectors, which completed successfully. This time, when installed in the Tivo, it started right up normally!!!

*****************************************************
My current (tentative) conclusions:

The mfsrestore process (not piped mfsbackup|mfsrestore copy) can fail when using the onboard raid connector to restore from a saved backup on the Soyo kt400 Dragon Platinum. Every time it hung at around 20%.

A piped mfsbackup|restore copy can be successfully completed using the Soyo kt400 Dragon Platinum motherboard when the non-raid ide connectors are used.

There could be an issue with the Iwill kk266r raid motherboard and mfsbackup. I was using a 1.3 athlon xp and 768 mb of memory.
I tried both standard backup/restore and piped backup|restore copy combinations on both normal and raid ide connections - with compressed and uncompressed backups. Either the decompression would fail, or there would be mfs_read_inode errors (in the case of uncompressed backups), or there would be no errors in the case of the piped mfsbackup|mfsrestore copies, but the drive wouldn't work anyway.

My last conclusion is that it would be a good idea to start a thread where likely problematic motherboards,raid connectors, & etc. would be listed to help the developers improve their product and to help the poor sots out who may be struggling with problematic configurations. Of course, maybe this already exists?
++++++++++++++++++++++++++++++++++++++++++++++

Anyway, I am a happy camper, although I camped for 2 1/2 days longer than I wanted to!!

Take care all -

smithcferg
smithcferg is offline Report Bad Post Report Post
Old 12-16-2005, 06:05 PM   #7 (Print)
HomeUser
Registered User
 
HomeUser's Avatar
 
Join Date: Jan 2003
Location: A2 Michigan
Posts: 529
I am not sure if a DMA problem is with MFSTools or the stripped down version of Linux on the boot CD not supporting the hardware in advanced mode.

Have any experience with 'C' ?
http://sourceforge.net/projects/mfstools/

Thanks for the feedback.
HomeUser is offline Report Bad Post Report Post
 
Forum Jump
Thread Tools

Go Back  TiVo Community Archive2 > Underground Playground > TiVo Upgrade Center

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off



All times are GMT -5. The time now is 12:46 AM.


Powered by: vBulletin Version 3.0.6
Copyright ©2000 - 2019, Jelsoft Enterprises Ltd.
(C)opyright - All Rights Reserved. No information may be posted elsewhere without written permission.
TiVoŽ is a registered trademark of TiVo Inc. This site is not affiliated with TiVo Inc.


Spider History Index