Segmentation Error, Drive stuck at Powering Up

[JJChandler]

Hopefully I can get some help since I've now failed at 2 drive upgrade attemps. I want to upgrade my 250G drive in the HR10-250 to a larger drive. I originally bought a 400G Seagate and followed the Hinsdale method. Everything worked fine. I installed the new drive and thought everything was working great. 2 weeks later the drive failed. (stuck on Welcoming, powering up... and I don't know why it rebooted in the first place.) I put the original drive back into the unit and it worked fine. I assumed the 400G was bad and returned the drive.

On my second attempt I bought a 500G Maxtor. I created a new backup file since I wasn't sure the backup file wasn't the problem. The backup creation took 45 minutes on my 486 machine. The restore took almost 3 hours? The Hinsdale guide says that the restore should be quicker. What's going on?

During my 3 hours I did lots of reading and I learned that I should have turned on DMA to speed up the process. I also learned that I should have used -r 4 switch in my previous mfsadd upgrade for larger drive support. Question 1) Is it possible my first attempt crashed after 2 weeks because of the missing -r 4 switch?

In my latest attempt everything fell apart in step 8 doing the restore. I used:

mfsrestore -s 127 -bzpi /mnt/dos/tivo.bak /dev/hdb

when the 3 hour restore completed I recieved an error:

Segmentation fault (which I read is memory related?)

I tried to install the drive anyway and the Tivo gets stuck in welcome, Powering up...

Question 2) Should I have used the -r 4 switch in the restore command? I thought that was only inserted in the mfsadd command??

Question 3) Can I reattempt this with the 500G drive? Would it make sense to use the original backup file that I used for first attempt only this time use the -r 4 switch?

Any other thoughts?

Thanks

[funtoupgrade]

mfsrestore -s 250 -r 4 -xzpi /mnt/dos/tivo.bak /dev/hdb for the 500GB drive and 200 instead of 250 for the 400GB.

[JJChandler]

funtoupgrade,

Thanks for the correction to my mfsrestore command. I went through the process again, got the segmentation fault and then got "Welcome, powering up..." loop. I read in some of your other posts that using the larger swap size of 250 required that I follow with a tpip command. I also read that the tpip command is only for the series 1?? Sorry but I'm confused and blurry eyed from reading these threads....

What tpip command should I use? I'm also using an old 486 machine, does that matter? The drive I bought is the Maxtor Ultra ATA/133 500GB 7200 RPMs. Is there any issue with this drive? I'm questioning everything now! Does the segmentation fault have anything to do with my problem? I read in other posts that this is a memory error and the upgrade worked despite this error.

Thanks for your help!

[funtoupgrade]

Oops! Forgot about tpip. Depends on which version you have. I have 1.1 and use:
tpip --swapped -s /dev/hd? where ? is new drive. Run right after completion of the mfsrestore. There is a different command if using tpip version 1.2 - if you need this one somebody will surely help or do a search on the forum.

For most series ones, you use copykern which is a tpip script that is generated as a result of user answering 6 questions.

Tpip needed if swap bigger than 127.

[JJChandler]

Well it looks like attempt #3 was a failure as well.

Here is the restore comand I used:

mfsrestore -s 250 -r 4 -xzpi /mnt/dos/tivo.bak /dev/hdb

followed by:

tpip -1 -s /dev/hdb

I received the following message after the tpip command: "initializing swap partion (version 1) @ 485512257, size=250Mib"

Tivo is again stuck in the "Welcome, powering up..." mode and has been like that for 15 minutes....how long should I leave it??

Here are the only odd things that happened during the upgrade:

- Received a Segmentation fault after the mfsrestore command. I have been just ignoring this error...is that the right thing to do?

- After creating the backup I noticed the following difference: Hinsdale says mfsbackup -f 9999 -6so /mnt/backup.bak /dev/hdc whereas the interactive guide says
mfsbackup -f 9999 -1so /mnt/backup.bak /dev/hdc What is the significance of the difference?

- The stock WD2500 drive has the jumper across pins 9 and 10. I can't for the life of me figure out what this is suppose to signify. According to the chart on the drive it isn't Master, Slave, CS or just parked. It works but I don't understand the jumper configuration. I've been setting the jumper drive on the replacement as Master.

Can this drive still be used or has it been corrupted by previous failed attempts?

Thanks for any help!

[JJChandler]

I am going to try this again this weekend on a different computer. Based on some reading it seems that people were able to get rid of the Segmentation fault by changing computers. Should I start from the beginning or just restore from the existing tivo.bak file. What is the difference between -6so and -1so? Hinsdale says to back up using -6so and the interactive guide says -1so.

Thanks,

[Freightdog]

My understanding of the -1so is:

-xso is for compression of the backup. x can range from 1 (minimal compression) to 9 (maximum compression).

6 (as in -6so) usually almost optimum compression on tivo types of data.

[JJChandler]

Thanks Freightdog!

Well it finally worked. I changed the computer I was using and everything worked fine the first try! I think my old 486 machine is toast. I'm going to try to copy all of my recordings to the new drive now. If it takes too long I'll just go back and restore from this working backup. Thanks for everyone's help!