Fixes for MFSTools 2.0 swap problems

[tfreitas]

mfsinfo /dev/hdb (A drive) gives me ...
Second MFS Drive Needed: No such file or directory
Second MFS Drive Needed2: Illegal seek
Second MFS Drive Needed: No such file or directory
Second MFS Drive Needed3: Illegal seek
mfs_load_volume_header: Total sectors(79108096) mismatch with volume header (313601024)
mfs_load_volume_header: Loading anyway.
mfs_load_zone_map: Primary zone map corrupt, loading backup.
mfs_load_zone_map: Secondary zone map corrupt, giving up.
mfs_load_zone_map: Zone map checksum error!

mfsinfo /dev/hdd (B drive) gives me ...
/dev/hdd10: Success
mfs_load_volume_header: mfsvol_read_data:Input/output error

mfsinfo /dev/hdb /dev/hdd gives me ...
mfs_load_zone_map: Primary zone map corrupt, loading backup.
mfs_load_zone_map: Secondary zone map corrupt, giving up.
mfs_load_zone_map: Zone map checksum error!
Segmentation fault

I can pdisk -l either drive, so linux sees the drives.

I had done an mfsadd earlier prior to having expanded the swap file. I then redid my dd_rescue so I'm now working with a fresh copy of the A drive. Maybe the B drive was effected by the earlier mfsadd? Any limitation on how many times you do an mfsadd?

Is it looking hopeless?

Thanks for your patience,

Tim

[tfreitas]

Here are the before and after partition tables, i.e., after the dd_rescue and then after I recreated it manually.

Original Table:
Partition map (with 512 byte blocks) on '/dev/hdb'
#: type name length base ( size )
1: Apple_partition_map Apple 63 @ 1
2: Image Bootstrap 1 1 @ 44023824
3: Image Kernel 1 8192 @ 44023825 ( 4.0M)
4: Ext2 Root 1 262144 @ 44032017 (128.0M)
5: Image Bootstrap 2 1 @ 44294161
6: Image Kernel 2 8192 @ 44294162 ( 4.0M)
7: Ext2 Root 2 262144 @ 44302354 (128.0M)
8: Swap Linux swap 131072 @ 44564498 ( 64.0M)
9: Ext2 /var 262144 @ 44695570 (128.0M)
10: MFS MFS application region 1048576 @ 44957714 (512.0M)
11: MFS MFS media region 32988398 @ 47054866 ( 15.7G)
12: MFS MFS application region 2 1048576 @ 46006290 (512.0M)
13: MFS MFS media region 2 44023760 @ 64 ( 21.0G)

Manually built table:
Partition map (with 512 byte blocks) on '/dev/hdb'
#: type name length base ( size )
1: Apple_partition_map Apple 63 @ 1
2: Image Bootstrap 1 1 @ 44023824
3: Image Kernel 1 8192 @ 44023825 ( 4.0M)
4: Ext2 Root 1 262144 @ 44032017 (128.0M)
5: Image Bootstrap 2 1 @ 44294161
6: Image Kernel 2 8192 @ 44294162 ( 4.0M)
7: Ext2 Root 2 262144 @ 44302354 (128.0M)
8: Swap Linux Swap 262144 @ 80043264 (128.0M)
9: Ext2 /var 262144 @ 44695570 (128.0M)
10: MFS MFS application region 1048576 @ 44957714 (512.0M)
11: MFS MFS media region 32988398 @ 47054866 ( 15.7G)
12: MFS MFS application region 2 1048576 @ 46006290 (512.0M)
13: MFS MFS media region 2 44023760 @ 64 ( 21.0G)
14: Swap Linux Swap 131072 @ 44564498 ( 64.0M)
15: Apple_Free Extra 79781120 @ 80305408 ( 38.0G)

[Robert S]

Yes, copying the A drive back on to the new drive after mfsadd'ing was a mistake.

I suppose forcing a GSOD might fix it (either boot with just the A drive attached or using Diagnostic mode (http://alt.org/wiki/index.php/TivoDiagnostics)), but I wouldn't be very optimistic.

[tfreitas]

So I'm giving up on preserving my recordings. I'd just like to get a working tivo. I did do a backup before all this started ...

mfsbackup -f 9999 -6so /mnt/dos/tivo.bak /dev/hdb /dev/hdd (A/B drives)

I just ran the restore ...

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

I get successful completion. mfsinfo looks good, telling me I have 2 more expands left and the increased time, etc.

But when I boot tivo, I'm still stuck in the boot loop. Got GSOD for a while then stuck in "powering on ..." and "almost there" loop.

I assume my image is good since I got succesful completion of all commands. Any suggestions? Swap is 127m.

Thanks again.

Tim

[Robert S]

That suggests a problem with the B drive. I would put an image on just the A drive and check that that works.

[tfreitas]

I already did that. Tried to boot off the A only first. Got GSOD then the loop. Then I mfsadd'ed the A to the B and got the same. Hmm. I wonder if my image is bad, though I thought I would have seen some indication of that during the backup, restore, mfsadd, mfsinfo, etc.

How about this. I have another tivo with a single 160GB. Can I backup that image and restore it on an 80 or 120GB - can an image be restored on a drive smaller than the one it came from? Or is it a function of how the partition table has been built regardless of whether I'm using the space? There's little to no recordings on the 160GB and I could clear the thing if that mattered. If I can't do that, any suggestions on getting a good image? Unless you think the problem is not with the image.

Also, would the first part of your Post#3 apply here (moving swap into inactive partition) or would that not apply since I have a 2 drive Tivo?

Thanks for all your help.

[Robert S]

You've got a 127Mb swap partition - that's the same size as your inactive root.

As long as the other TiVo is the same model (the first three numbers of the service number must match), then there's no problem taking a backup from that one.

[tfreitas]

My bad tivo is an HDVR2 (1st 3 digits 151). My new tivo is an SD-DVR40 (1st 3 digits 351). Both Hughes - the SD-DVR40 is the newer version of the HDVR2. Am I out of luck?

[Robert S]

Er, yes, I think that should be OK.

You might want to read the 'Moron' artical at tivo.samba.org, so you'll know what to look for.

[tfreitas]

Hmm. So is that your subtle way of calling me a moron for asking a dumb question or do you actually think it has a chance of working but I should be aware of possible consequences as outlined in the moron thread. I'd like to think the latter :). I'm thinking I may go to Instantcake and get a $20 virgin image for an HDVR2.
Tim

[Robert S]

Is the difference between 'Moron' and 'moron' too subtle for you?

Like I said, I think you'll be OK - the Series 2 DTiVoes seem to cope with the Type II Moron situation better than other models and will download and successfully install the correct version.

But I would still check that this does infact happen and stop the TiVo phoning home if it doesn't. (At that point IC starts to look like an attractive proposition).

[tfreitas]

Kudos to Instant Cake. For $20, I get a virgin image and a working Tivo. Just have to "clear and delete all" to get things up. Thanks for all you help Robert S. I think the problems I had recovering the system all stemmed from the fact that my backup image was taken after I already had a harddrive problem. So it was all for naut. Though I learned a lot and will be in better shape next time I inevitably have HD problems.
Thanks again,
Tim

[richard27]

I am trying to add a 200GB HD to my RCA DirecTV Tivo (has a 40 GB in it already). I have followed all of the Hinsdale steps until step 8. The last command I used was:

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

According to the steps, since my new drive is over 140 GB, I am supposed to use one of the following commands next:

mfsbackup -Tao - /dev/hdc | mfsrestore -s 127 -xzpi - /dev/hda
or
mfsbackup -Tao - /dev/hdc | mfsrestore -s 127 -xzpi - /dev/hda /dev/hdb


Am I doing this correctly? Which one of these do I use? The second one assumes that I have 3 drives, but I want to keep my original tivo drive. Can I just use the mfsadd command? Also, after this step, are they ready to go back into the TIVO?

Sorry for all the questions, but I am a little lost.

[Robert S]

You can't increase your swap if you keep your original A drive. You could to the first pipe to copy the recordings to the new drive and then mfsadd the A drive as a B drive.

However, you don't actually need more swap at this point as your total drive capacity will be 177Gb, which is known to be OK on a DTiVo.

Having said that, if you're thinking of upgrading further in future, it's a lot easier to expand swap now than on subsequent upgrades.

[Scot]

I've got two pairs of 120 GB drives for a Philips Model: DSR6000R01, and twice now I've run into the GSOD problem when the drives fill up with files I've set to "Keep until I delete". The only way I've found to get things going again is to treat the drives as empty and re-do the upgrade procedure, including increasing swap space using the s -127 option in the mfsrestore command.

Is there any way to save the data on my drives? I've let them "cook" in the TiVo overnight, hoping the GSOD repair routines would perform a miracle, but so far no joy. Mounting them in a PC and booting off the MFSTools disk wouldn't allow me to cruise the directories on the TiVo drives. My though there was that perhaps deleting some data files would let the drives work again?

Before I wipe out this set of drives and re-upgrade them, what steps can I take to make sure that the GSOD doesn't happen again? I'm following the Hinsdale instructions, upgrading two small drives to the two 120 Gb drives.

Help!

TIA,

Scot

[SEC55]

I just completed the upgrade to my stock Sony SVR-2000. I'm pleased to report that I now have 194 hours! I'm quite happy that I did it myself, because I'm not much for command lines and jumpers.

However, I did run into one problem. I was unable to make MFS Tools recognize the s -127 command to increase the swap file, even though I followed the Hinsdale How-To exactly. So I just did the recording capacity upgrade and I have some questions:

Is there a way to go back and increase the file? If so, how?
Does it matter?

I looked over this forum - although I didn't read all 26 pages! - and saw one post that seemed to say it wasn't that important, and others that said it was.

BTW, I unknowingly booted into Win2000 before reading (at the very END of the tutorial) that I shouldn't do that. That info should have been one of the first bits of advice given. Both drives seem to work, though.

[Robert S]

"s -127" or "-s 127"?

Anyway, it's a bit late to do anything now, unless you want to repeat the upgrade.

Print out the third post of this thread and keep it in a safe place.

[SEC55]

Originally posted by Robert S
"s -127" or "-s 127"?

Anyway, it's a bit late to do anything now, unless you want to repeat the upgrade.
Could I just repeat the capacity upgrade part, or would I have to start from scratch?

[Robert S]

You'd have to to the whole thing over.

[SEC55]

Well, I supposed I could redo it if it's really necessary. Two questions, however,

1. I didn't back up my recordgings, just did the capacity upgrade, so I used the line

mfsadd -x /dev/hdc /dev/hdb

to do the upgrade. Where would I add the 127 command? I didn't know where to put it and it wouldn't go.

2. Would it be better to use my new 7200 rpm drive as the A drive instead of the stock 5400 rpm?

[Robert S]

-s <number> is a parameter for mfsrestore, not mfsadd.

The swap partition is on the A drive. There's no way to affect the size of the swap allocation without altering the A drive.

If you'd wanted to increase swap, you'd have had to copy your recordings to the new drive and then (optionally) added the old A drive as a B drive.

[Snorwegian]

Sorry to keep asking essentially the same question Robert, but I just wanted to make sure-- I pored over this thread but couldn't find a specific response.

I have a Series 2 Samsug DirecTivo with a 100GB HD already in it (factory installed). I want to add a 120 GB HD as a B drive to it. As I understand it, I will NOT be able to increase the swap file size on my A drive without losing all my existing recordings (unless I transfer my 'A' drive to my new HD, and then using my A drive as a new B drive). Is this in face the case ?

If so, then my only real option is to simply perform the capacity upgrade without increasing swap file, and in the case of a GSOD, then I can use your instructions from the third post to try to save myself, correct? But in that case, I'd lose all my recordings anyway right?


Thanks again for all your help; you guys are the best!

[Robert S]

That's correct, except for the last line: The point of temporarily increasing swap is to SAVE your recordings by allowing the GSOD to complete. If you're happy to lose your recordings, you'd just reimage from a backup.

[Scanman0]

I origionally added and expanded a 250 GIG a drive, before MFStools 2 then used MFStools 2 to add a second 250 GIG B drive to obtain a cool half a terabyte on my series one hdr312 tivo.
My swap situaion is:


Memory Statistics:
total: used: free: shared: buffers: cached:
Mem: 14274560 14131200 143360 65064960 49152 4120576
Swap: 314564608 1765376 312799232
MemTotal: 13940 kB
MemFree: 140 kB
MemShared: 63540 kB
Buffers: 48 kB
Cached: 4024 kB
SwapTotal: 307192 kB
SwapFree: 305468 kB

It's filling up fast, as I made "Best Quality" a nice DVD compatable recording resolution :D

Also running cashecard w/512 stick.

Has anyone ever GSOD from a true MFS corruption and recovered with 500 gig's of stuff on a series 1 standalone ? If so how many DAYS did it take ?

[diskus]

so let me ask a question, if I am upgarading a tivo to a 120 and 200gig drive combo giving 320 total, then there is no reason to increase the swap file as it is over the maximum either way? ( 276 i believe was the figure in a earlier post)

is this correct?

thanks

MB

[Robert S]

274, actually.

If you're replacing the kernel, then you get the ability to use swap partitions larger than 128Mb, so you can get enough swap for the GSOD to run on really large drive sets.

[HTH]

I've just recovered from my first GSOD-reboot loop on my oldest upgraded TiVo. Hit it while recording an unexepcted (new?) episode of Full Metal Challenge.

Am I right that the only permanent solution to an insufficiently large swap is to either (a) upgrade to a larger primary drive if possible or (b) lose all recordings?

I'm already using two 120 GB drives in a Series1 and upgrading to a newer kernel to fully utilize a larger primary drive sounds just as risky on an upgrade (if one ever happens) as leaving my swap pointed at hda7.

Is there anything unsafe in having a partition in >137 GB space referenced in the partition table that can only be accessed when the drive is hooked up to a PC, say holding scratch backup space?

[Robert S]

You think you'll get a software upgrade before you'll have to replace those drives?

I think TiVo have completely ruled out further software upgrades for Series 1's, so, assuming this is a stand alone model, I don't think you need to worry about that. (DTV seem to be continuing to tweak the Series 1 DTiVoes).

So, you might as well leave the swap on hda7 for now.

I suppose it might be possible to steal some space from hda7 and/or hda9. Both are contiguous with the swap partition and neither is full. It would be such a fiddly task, though, that it's probably not worth contemplating.

We have seen TiVoes with LBA-48 images and non-LBA-48 kernels. They only misbehave when they start to fill up, suggesting that the kernel only tries to look beyond the 137Gb mark when it needs to record something there, so I don't think your partitioning idea would cause a problem.

[scoombs]

I added a 100GB B-drive to my 30GB S1 almost 3 years ago. I just combined them to a 160GB replacement A-drive by using this command from the Hinsdale How-to:
mfsbackup -Tao - /dev/hda /dev/hdb | mfsrestore -s 127 -xzpi - /dev/hdc

Since I am under the 140GB trigger, does that drive have only a 64MB swap, or was it expanded because the -s 127 option was used for the mfsrestore? That leads to my real question, that if I add a 100GB B-drive back to the system, taking the total size to 237GB, am I good to go swap-wise with just doing the following?:
mfsadd -x /dev/hdc /dev/hdb

Reading the response to twash on 7/7/04 suggests I am good to go since the A-drive image was created using the -s 127 option, but I am just trying to be cautious. :)

Thanks much

[HTH]

Originally posted by Robert S
You think you'll get a software upgrade before you'll have to replace those drives? With TiVo's future plans in regards to premium and PPV content, I worry that their partners may demand it get an upgrade. It's a small concern, but a concern nonetheless. (I don't want to discuss that topic in this thread.)

It is a standalone.

I suppose it might be possible to steal some space from hda7 and/or hda9. Both are contiguous with the swap partition and neither is full. It would be such a fiddly task, though, that it's probably not worth contemplating. I seem to remember way back when it was said that updated kernel and root partitions were simply dd'd into place on an upgrade. If that is the case, it may be unwise to try to steal space from another partition.

We have seen TiVoes with LBA-48 images and non-LBA-48 kernels. They only misbehave when they start to fill up, suggesting that the kernel only tries to look beyond the 137Gb mark when it needs to record something there, so I don't think your partitioning idea would cause a problem. That settles some concerns, though I doubt tests with partitions wholly inside the LBA-48 area have been been tested. I'll consider giving it a try and will report any unexpected results when time allows.

[n00blarME]

Help anyone? After attempted upgrade using Hindsales How-to and PTVMFStools2 iso and weaknees bracket/powetrip/fan, my TiVo is stuck through Powering Up to Almost There then the GSOD! Pleaseeee im dying here!!!!

[djliquidice]

What happens if i omit the swap space parm?

[Robert S]

You get 64Mb of swap

[djliquidice]

Thanks. :)

[janol]

Hi Robert,

I hope you'll be able to help me in my situation.

I currently have a Series 2 TiVo with 2 drives in it: the original 40GB drive and the one I added about a year ago 120GB. When I was doing the upgrade, I chose the option to just do mfsadd command and only later found out that I may have a problem in the future because the swap space was not increased.

Now I have a problem. My TiVo has not GSOD'd yet, but it started experiencing "hick-ups" in video and audio and I decided to check my drives for errors. I used Maxtor PowerMax and it indicated that the original 40GB drive is failing.

I want to move my data from the bad drive into a new one. The new one will probably be bigger. Is there any way for me to get the data out of my bad drive and increase the swap size in the process?

Any help will be greatly appreciated.

[Robert S]

There's no immediate problem with swap. Series 2 Tivoes have more RAM than Series 1 stand alones, so you can fit up to 180Gb without requiring extra swap.

However, that upgrade will take you over the limit.

The third post of this thread describes what to do. dd the drive on to the new one, use pdisk to create a new swap partition, initialise it and then mfsadd.

[janol]

Thanks Robert. I am going to give it a try. Just a couple more questions.

A while back in the posts user 'tfreitas' seamed to have the same problem I do right now and you said that this whole thing may not work because he did mfsadd during initial upgrade.

If I did mfsadd during initial upgrade will I still be able to 'marry' my new drive A and the old drive B making sure I do not loose my recordings?

Also, what part of the post #3 I should use? It seams the situation on the bottom is the one I am facing now.

Thanks.

[janol]

I tried dd_rescue the drive today. Got a million errors and decided that that it would probably be more safe to just forget the recordings and restore the backup that I made during my original upgrade.

Are the following commands proper to restore the backup onto a new A drive?

mfsrestore -s 127 -xzbpi /mnt/dos/tivo.bak /dev/hdc

Assuming new drive is Secondary Master and Windows C: drive is Primary Master.

Do I have to execute the following commands before the restore to mound DOS partition?

mkdir /mnt/dos
mount /dev/hda1 /mnt/dos

After the restore is done, do I have to run

mfsadd -x /dev/hdc /dev/hdb

to marry the old B drive to the new A drive (assuming old B drive is Primary Slave)

Old B drive has data on it. Will I be able to see at least some of the recordings that are located completely on the B drive or will they be gone? Is it a good idea to leave the data on this drive at all or should I just restore onto both drives at the same time?

Thanks for all your help.

[Robert S]

You have to mount the DOS partition so you can read the backup

You can (and should) restore to both drives

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

None of your recordings will survive this, mfsrestore will wipe everything out.

[janol]

Thanks Robert,

I understand that I should restore to both drives, I am just curious, what would happen if I just restore to the new A drive and mfsadd the old B drive?

[mbellot]

Just wanted to give a big thanks to Robert and everyone else who have already "been there, done that".

My Tivo went into an endless reboot loop (green screen "severe error") earlier this week and I thought I was screwed.

After digging through the forums I found this MFSTools bug thread, although I don't think that was strictly my problem...

I recently upgraded my HDR312 to a 250GB drive (using the PTV LBA48 disk), and I had (foolishly) thought if 127MB of swap space was good that even more would be better, so I told MFSTools to reserve 300MB of swap space.

According to the instructions the swap space would be initialized when I installed the new kernel but...

After doing some reading it appears the S1 can't cope with any swap space larger than 127MB, so I effectively had nothing.

I yanked the drive out and installed it in my PC, boot the PTV CD and try to mount the /var partition (just to be sure I got my byte swapping right). Mount returned an I/O error. A quick check with pdisk confirmed that the partition table was intact (meaning byte swapping should be correct), so now I'm thinking the drive is toast.

I go and download the PowerMax utility (its a maxtor 250gb drive) and find out its only reporting 9MB of drive space!!!

More forum searching turns up the fact that the drive is probably locked and that I need to use diskutil (and NOT qunlock), so I unlock the drive and reboot into the PTV CD. This time everything goes smooth with pdisk and mounting /var so I run mkswap -v0 /dev/hdc8 130048. BTW - The PTV CD did NOT unlock the drive even though several posts claim the various boot CDs should.

Put the drive back in the Tivo boots to the green screen (severe error), but does NOT reboot (woohoo!). Fifteen minutes later the machine reboots and I'm back online. It even appears that all my recordings survived this nasty little ordeal.

So, for anyone who doesn't believe that most of the answers are already here, think again. I did everything necessary to fix the problem without asking a single question.

Again, a world of thanks to those who have already figured this out.

:up:

So now I have two questions :D

1. Can anyone definitely confirm that an HDR312 will not work with more than 127MB of swap space.
2. Assuming the answer to #1 is yes (127MB limit is real) Is there any way to utilize the remaining space for other things (like tivoweb, etc) so I'm not loading up the root and var partitions?

[Robert S]

You should have used tpip instead of mkswap to initialise the swap.

mkswap is limited to 128Mb, although it clearly shouldn't be.

However, you have enough swap for at least 274Gb of disk space, so you're fine for now.

But if you add a B drive, run tpip at that point.

[mbellot]

You should have used tpip instead of mkswap to initialise the swap.

mkswap is limited to 128Mb, although it clearly shouldn't be.

However, you have enough swap for at least 274Gb of disk space, so you're fine for now.

But if you add a B drive, run tpip at that point.


Thats the odd thing.. According to PTV (I used the PTV LBA48 CD image)

CopyKern will automatically initialize your swap partition, so the aforementioned tpip command would be unnecessary when using CopyKern.

So when I used copykern to install the LBA48 kernel I assumed it was initializing the swap space.

I don't know what happened, but obviously the swap space did not get initialized.

[ronsch]

Robert,

I just got a used DSR6000R01 sans drive and took a backup image from a friend with the same model. I used mfs tools 2.0 to restore the image to a 160gb drive without expansion (-zpi) and all seemed to go well. Putting the drive in TiVo and powering up, I get an endless reboot loop . In all likelyhood, the original backup was made using mfs tools 1.1. Could this cause this problem or is something else wrong with this TiVo?

Update: The backup was taken two years ago so it probably was done using mfs tools 2.0.

Update : I unplugged and replugged and the unit no longer reboots. Just sits at Welcome Powering Up with regular disk activity.

Update: rejumpered the drive to Master instead of CS and replaced ide cable with a brand new one. No change.

[TomTomaro]

I used the hinsdale guide to upgrade to a larger disk and now get the power up screen and nothing more, using either the replacement or original drive. Any suggestions?

[rkcarter]

Can you hear the drive spin up? A friend once managed to catch me just in time from closing up the box without reconnecting the power supply to my new drive, for example.

[TomTomaro]

Great first question. Simple stuff first. Yes I can hear the drive spin.

[MannyVjr]

OK, here is where I get a little confused...

D. Renumber the partitions.

pdisk /dev/hda

On TiVoMad, pdisk is /mad/pdiska.

If your inactive root partition is 4, type:

r 8 4
r 5 9

If your inactive root partition is 7, type:

r 8 7

w (write partition table)
y (confirm, returns you to command prompt)

E. Restore drive to TiVo

Let the TiVo run, it may take a long time, and may reboot occasionally, but it should eventually boot up properly. Remember, you must complete this sequence, your TiVo is not safe when it recovers from the green screen.

F. Restore the original partition numbers

Repeat step D above.

You can now safely replace the drive in the TiVo.


On step E it states to replace the drive on the tivo, so as soon as I get the command prompt (from step D) I power down my PC and extract the drive, right?

...and I'm curious when you say it might take a long time (the TiVo running to fix itself), you mean 2-4hrs or more like 4-6hrs?

[slaponte]

Would it be too much to ask if somebody can resume the 19 pages of thread into an answer? <Edited to answer my own question>

A) When do you need extra swap? If total disk space will be 128Gb or more, you need more swap. If you don't, under certain conditions, your Tivo might crash and you will most likely loose all your recordings.

B) How much do you need? You need 1MB per every 2GB of total disk.

C) Sequence of commands to add the swap and or fix headers?

Use mfsrestore flag "-s" to set swap size. Up to a value of 127 it will work fine. So that covers up to 254GB total disk. If your total disk will be larger than 254Gb, then you need to

- specify new size on the mfsrestore. (Example, "-s 200" creates 200MB swap for a 400Gb disk).

- run TPIP to initialize the swap header

tpip --version shows the version number you have

tpip V1.1 : tpip --swapped -s /dev/hd?

tpip V1.2 : tpip -1 -s /dev/hd?

Tpip can be downloaded from the web (V1.2) and is also found (v1.1) in the PTV Upgrade ISO CD image.

[Bunny]

I posted this in another thread but this thread may be more appropriate... I have an hr10-250 which I upgraded a couple of months ago with 2 seagate 400gb drives. Unfortunately, at the time the whole swap file discussion had not started yet. As far as I know, I have a 127mb swap file, whereas I understand it shoul be 400mb. I have not seen the GSOD but I do suffer from the occasional stutter, which I think may be related to the swap file size. I can redo the whole procedure again but I rather not want to lose my recordings and I do not have a couple of spare masssive drives to backup to. Is there any other way to do this without doing backup and restore?

Thanks!

[slaponte]

That I know off, no. To change the size of the swap space you have to modify the file system and that requires a backup, then restore with new structure.

I have read (deep in this long thread) a way to use an unused partition in the file system and add it as swap. But I seriously doubt it will give you the extra 270Mb+ you need.

The low swap won't matter until you hit a GSOD that loops and loops... and then, you will probably loose anything you have recorded.

[tonsa]

I upgraded mine to one 160gig and one 320gig but when I was done at first with 1 drive upgrade it showed I had 200 hours instead of the 160 hours then I added another drive it added 380 hours instead of 320 hours so all in all I have 580 hours which is wrong I just added 480gig it should show 480 hours right? is there something wrong to what I am doing? I followed everything.

Is there a way to upgrade without adding the original drive size to the upgraded drive? even though the original drive is removed?

I am confused. I use weeknees upgrade and then I tried using msftools 2.0 still the same.

someone help? I am using 2 western digital drives

thanks in advance

[mvita]

What is the proper tpip command to use to initialize increased swap space on a HR10-250? Earlier in this thread, the syntax (for tpip v1.2) is shown as:

tpip -1 -s /dev/hd?
I'm a bit confused about the "-1" argument... the tpip doc says that this means "Treat the disk as a Series 1 TiVo disk instead of letting tpip figure it out". But isn't the HR10-250 essentially a "Series2" unit? Should I be using "-2" instead?

[slaponte]

No. For some reason the swap header from the S1 will work.

I can't explain why. Only know that everybody who used it got it working (on S2 machines).

[slaponte]

Tonsa, you don't explain exactly how you are doing this, so it is hard to tell what you are doing wrong... :)

[mvita]

Okey doke. Just wanted to make sure I wasn't screwing things up.

Thanks....

[tonsa]

Tonsa, you don't explain exactly how you are doing this, so it is hard to tell what you are doing wrong... :)

Here is what did both on weekness, hinsdale and the msftools

mfsbackup -f 9999 -1so /mnt/backup.bak /dev/hdc which is the b is the original tivo

then I did this

mfsrestore -s 127 -r 4 -zxpi /mnt/backup.bak /dev/hdZ /dev/hdZZ

which is the Z is the new master drive 160gig and the ZZ is the new slave drive 320gig

it still gives me extra over 40 something gig extra on each hard drive.

Is this correct?

Thanks

[azitnay]

On standalone units, the hours-to-GB ratio is slightly more than 1-to-1. This is less evident on the early Series2 models, where 40, 60, and 80GB were marketed as 40, 60, and 80 hours (although in reality they could record a little more than that), but on newer models such as the 140-hour models, they're really running 120GB drives. The biggest TiVo on the market, the Humax T2500, uses a 250GB hard drive and advertises 300 hours.

Personally, both my units with 160GB show ~180 hours, and my unit with 120GB shows ~140 hours.

Drew

[tonsa]

Thanks alot so I have alot of hours now! I am glad its clear now. You guys are very very helpfull! I really appreciate it.

More power to you guys

[lennonc@gmail.co]

I cannot find the mkswap file that is need to perform the swap increase mkswap -v1 /dev/hdc

[slaponte]

mkswap is a Linux command. It would usualy be in /bin, /sbin, /usr/bin or /usr/sbin

A way to find it

cd /
find . -print | grep mkswap

This will return to you the full path to the command.

[GGTZ]

I have installed a 160 GB drive sucessfully into my series 2 Tivo and it's working great. I'd now like to add my orginal 40 GB drive as drive B. I probably should have done this when I did the upgrade to the 160 GB drive but I didnt.

Can I now add this as drive B (the orginal 40 GB drive) after I have expanded drive A to 160 GB.

Basically... I've now got a TIVO with a 160GB single drive TIVO working fine. But, now I've got the 40GB drive sitting doing nothing so I thought I'd go ahead and add it.

What commands do I use...or that is instructions in the Hinsdale upgrade?

The instructions for ugrade configuration #1 are indicating and "unmodifed or Expanded".

[slaponte]

I am not sure it is worth the effort. You would have to backup the 160 back to a "shrunk" image to come back to the 160+40 and probably loose any recordings. You can't expand an expanded drive. Or you can start over with the 40Gb image and go forward from there. In that case is easy, just restore the 40GB backup image into the "160+40" (note that you will need to set swap at 100)

Many people subscribe to the idea also that you are adding more chance of drive failure with two drives. I am not usualy very concerned about that, but hey, just info for you.

[rkcarter]

IMHO you may want to keep the 40 as a nice safe backup of your TiVo's original configuration, which will come in handy to throw in to see if a problem is a TiVo problem or a disk problem; also if it's still under any sort of warranty you'll want the original disk back into it I think before you take it in.

[2devnull]

In that case is easy, just restore the 40GB backup image into the "160+40" (note that you will need to set swap at 100)

Many people subscribe to the idea also that you are adding more chance of drive failure with two drives. I am not usualy very concerned about that, but hey, just info for you.


If I backed up my original Tivo 80 GB drive and then want to restore it to both that original drive (80 GB) and a new 320GB drive, what will be the tpip command I would need to run using tpip 1.2?

tpip -1 -s /dev/????

Which drive should the tpip command be run on?

[2devnull]

Here is what did both on weekness, hinsdale and the msftools

mfsbackup -f 9999 -1so /mnt/backup.bak /dev/hdc which is the b is the original tivo

then I did this

mfsrestore -s 127 -r 4 -zxpi /mnt/backup.bak /dev/hdZ /dev/hdZZ

which is the Z is the new master drive 160gig and the ZZ is the new slave drive 320gig

it still gives me extra over 40 something gig extra on each hard drive.

Is this correct?

Thanks



you only have 127 for swap? Is this working out for you?

[azitnay]

If I backed up my original Tivo 80 GB drive and then want to restore it to both that original drive (80 GB) and a new 320GB drive, what will be the tpip command I would need to run using tpip 1.2?

tpip -1 -s /dev/????

Which drive should the tpip command be run on?

Whichever drive you make the primary (A) drive.

However, as an aside, I'd recommend just using the 320, and leaving the 80 out of the equation.

Drew

[2devnull]

Well, I ended up restoring with 256 MB swap to the new 320 GB drive, running tpip on it, then adding the original 80 GB drive as a B drive. Got 350 hrs now.

Everything seems to be working correct so far. Now sure why everyone is recommending against adding the orig. 80 GB drive as a B drive.

[JamieP]

Now sure why everyone is recommending against adding the orig. 80 GB drive as a B drive.A two drive system has roughly double the failure rate of a single drive system. If either drive fails, you lose everything -- all recordings, etc, and will have to recover from a backup image. It might be worth it for two 400GB drives in an HDtivo, but most of us think it isn't worth it to add a small drive to a much larger one.

[2devnull]

does anyone know the reason a two drive system failure rate is double? Is it cooling, processing power, bus limit etc?

As far as I know, those at ptv andweeknees who do a lot of these two drive upgrades have not agreed this to be an issue.

[JamieP]

does anyone know the reason a two drive system failure rate is double? Is it cooling, processing power, bus limit etc?It derives from basic probablity. See the additivity section here (http://en.wikipedia.org/wiki/Mean_time_between_failures).

[2devnull]

What really does not make sense is that by doing two 400 GB drives, you are going to loose more since that does not change the risk of the drives failing but does potentially risk you loosing more of your recordings due to the bigger combined size.

I conclude this to be a non-issue unless there is something more substantial that would indicate why adding a second drive would actually increase risk of failure.

If we worry about the laws of probability, the world would still be primitive.

[azitnay]

It's not a huge issue, and I still have one two-drive TiVo running... But the bottom line remains that if you have two hard drives, each with its own chance of failing, you get roughly double the chance of at least one of them failing. Once that happens, as JamieP said, you're usually out of luck with respect to everything on the drives, since they're a married pair.

Drew

[JamieP]

I conclude this to be a non-issue unless there is something more substantial that would indicate why adding a second drive would actually increase risk of failure.To put this into perspective, a typical disk has a MTBF of 1 million hours. That works out to be something like 1 failure every 100 years. With two disks the failure rate doubles to 1 failure every 50 years. That increased risk might be a non issue for you, and there is nothing wrong with that. I'm just trying to explain the rationale for the original advice you were given.

If we worry about the laws of probability, the world would still be primitive.Ignore probabilty at your own risk. :-)

[2devnull]

Well the thinking should be that adding a second drive just gives more space and no one should be assuming what is store there is recoverable or is separate. Basically, you view your system as a one drive system in either case.

Now, other than additivity theory, there seems to be nothing conclusive that states that adding a second drive will make failure more likely due to a technical reason. Even with a one drive system, you still risk that single drive failing in any event.

I think the whole probability thing is blown out of proportion and mis-understood. Drives these days with the proper care are not likely to fail routinely.

[2devnull]

BTW - JamieP and Drew, Thanks for the discussions. Don't mean this to sound arguementative. Just wanted to find any technical reasons behind the risk that can be addressed/mitigated potentially with some modding. Thank you.

[rkcarter]

It won't double the overall failure rate of the TiVo, but drives seem to be the most (or maybe 2nd most behind modems on TiVo 1's but I dunno about later TiVos) failure-prone item in a TiVo. So you're putting in two of the most failure-prone item, doubling the chances of a drive failure. And of course if the modem fails, you don't lose all your recordings, and have a few days of schedule to replace it with an external modem (been there done that).

[rkcarter]

Oh, and you are adding heat and power consumption, so I don't know how significantly or insignificantly but that will put more strain on other components.

[2devnull]

From what I have read on weaknees, adding a second drive actually makes the system run coler since an extra fan(s) is added.

[rkcarter]

Interesting ... when I upgraded my TiVo 1, there was nothing in the instructions about adding a fan. Do you have to cut a new ventilation hole in a TiVo 2 for the fan?

[azitnay]

Assuming your post wasn't meant to be sarcastic, 2devnull is probably referring to systems whose upgrade included a WeaKnees TwinBreeze bracket, which adds an internal fan among other things. Certainly not a "standard" part of any upgrade, but WeaKnees does indeed make the claim that a two-drive system with a TwinBreeze runs cooler than a one-drive system with the stock fan and bracket.

Drew

[2devnull]

Drew is absolutely correct. I am using the new G2 version of the Weeknees twinbreeze.

[chelman]

My computer has a SATA drive and MFSTools gets stuck while booting. What can I do?

[whatever]

My computer has a SATA drive and MFSTools gets stuck while booting. What can I do?I think you need to be more specific on "gets stuck while booting". Are there any error messages?

I haven't tried booting mfstools (yet), but since it's just a stripped down linux distribution, I suspect that there is no SATA support in it. Typically, SATA drives are detected using SCSI emulation with linux distros. I suspect that mfstools doesn't have any SATA support built in.

You might want to try downloading a "live" linux distibution and booting to that. It will boot from a CD and not touch your existing SATA drive. You could then run mfstools from a floppy.

[slaponte]

It's been a while. Hi everybody!

2dev, I think the statement many make of "increased chance of failure" with two drives is missleading (one of those "you can make statistics say anything you want" things.).

The phisical fact of having two drives does not increse the chance of either one of them failing, it increases the chance that the Tivo would have a disk failure (with 2 devices you have double the chance that one might fail at any given time). Similar to this : if you don't have power windows on your car you have 0% chance of them breaking, if you add power windows you now have double the chance than before...

The problem is that if you loose one disk on a dual-disk tivo you might/will loose everything you had recorded. Now that to me is not a big deal, so I discounted the risk. If a loose a disk, well, too bad, I get to install another one and reformat and go on with my life. Statistically speaking, I had 0% of those recordings before the Tivo, so I am no worse than when I started if I loose them ;) ;)

[2devnull]

yep. fully agree. Whether you loose the single drive or the dual drive (either drive), the net result is the same thing, you loose all your recordings. And going from 1 drive to 2 does not contribute to a failure is the point needed to be made.

[rkcarter]

It's been a while. Hi everybody!

2dev, I think the statement many make of "increased chance of failure" with two drives is missleading (one of those "you can make statistics say anything you want" things.).

The phisical fact of having two drives does not increse the chance of either one of them failing, it increases the chance that the Tivo would have a disk failure (with 2 devices you have double the chance that one might fail at any given time). Similar to this : if you don't have power windows on your car you have 0% chance of them breaking, if you add power windows you now have double the chance than before...

The problem is that if you loose one disk on a dual-disk tivo you might/will loose everything you had recorded. Now that to me is not a big deal, so I discounted the risk. If a loose a disk, well, too bad, I get to install another one and reformat and go on with my life. Statistically speaking, I had 0% of those recordings before the Tivo, so I am no worse than when I started if I loose them ;) ;)

As one of the ones who brought up "increased chance of failure," I want to make clear that I meant EXACTLY as you said here. I certainly did not mean to say nor imply anything other than that -- 2 pieces have a greater error rate than one of the same pieces, BECAUSE there are two of them that can break, and in this case, either of them breaking causes you to lose the recordings.

I had about the same number of recordings pre-TiVo -- just more videotapes. ;)

[slaponte]

I have one dual and one single. Today, my advice would be stay single.

- easier upgrado to do
- less noise and hassle (power splitter, fans, etc)
- I definetly would not miss the 40 or 80GB when you have 300Gb in there.

You DO double the chances of loosing the Tivo.

I recently re-used one of my 40Gb drives on a PC, and the 80Gb went into this nice USB enclosures they make now. You drop the IDE drive in it, and it turns into a portable USB 2.0 device. Very cool way to add some backup space to your PC.

[Robertjm]

Is this still a problem with the latest versions of MFStools, or has that been fixed?

[slaponte]

I am not aware of a newer version of MFSTools. By the threads it seems there is a recent version of MFS tools. By the threads it would seem there is a "new" version but if you pay attention to the dates that was a couple of years ago...

But I have been away from these forums a while, I could be wrong.

[kevincol]

I have a HR10-250 that I have had for years.

There was a power surge and my unit started to lock up.

I checked drives with SpinRite and they are both fine.

Determined that there must have been corruption on OS files.

I wanted to go to "F" release and bought InstantCake and installed "F" version.

Machine booted up fine and seemed to work. However, it was only recognizing one drive in terms of space (i.e. 30hr HD on system info).

I removed second drive and machine still booted, indicating that it wasn't really seeing second drive, though TivoWebPlus Info shows both drives present.

I took drives out and used MFSAdd to pair the drives back together. MFSAdd said it added another 300hrs of SD recording time (i.e. from 281 to 581). Did [CTRL] [ALT] [DEL] to close down session and put drives back in Tivo.

Tivo still reports 30hours HD and doesn't report drives as combined.

Pulled drives out again and used MFSAdd to bring both drives down to one drive. MFSAdd reported drive down to 281 SD hours. [CTRL] [ALT] [DEL] to close down session, come back into MFS Tools 2.0 and use MFSAdd to pair drives again. MFSAdd reports drives were paired and that I should have 581 hrs SD.

Put drives back in Tivo and still reports 30hr HD record time!

Why on earth won't the HR10-250 see the entire space.

Is it because they were paired under the baseline OS and upgrading to "F" OS release is causing problems?

I'm at a lose and outside of FDISKing the second drive, dont' know what else to try.

[Robertjm]

Recently I upgraded my Series 2 540 with a WDC300GB drive. I had problems when I tired creating it with -s 200 for the swapfile so I went ahead and used -s 127. The upgrade seems to have worked just fine (so far) and I haven't had any glitches (so far).

After pawing around the Tivo Underground archives I have read multiple mentions of dropping into the Green Screen of Death and basically losing everything on the drive, and the size swapfile I created was good for up to 274gb drives. Is there a way to upgrade the swapfile with all the new recordings that I've done staying intact, or should archive them to my PC first? Since 274gb is so close to 300, and I more often than not delete the recordings after watching them once, do I really need to worry about it? I have yet to get anywhere near to filling the drive up, but I expect to be putting a lot more on the drive though once the Olympics roll around this coming Friday.

Thoughts?

Robert

[proxis]

Hello All:

Well? I need help from on of you kind gentlemen. I'm posting here because I believe my problem is swap related.

I have a Humax T800 whose HD was starting to die (freezing, skipping). I am an IT consultant and have no problem handling hardware and Windows based stuff, but I don't know anything about LINUX, not yet anyway. (some UNIX experience though).

So I bought a new drive, a DAM10 (one of the latest ones without the problems) and I also bought the PVT Tools Universal CD because it had MFSTOOLS 2.0 and EVERYTHING on it, supposedly.

So I decided I was going to use Hinsdale's Option #3 for Single - to New Single, larger, - this is the slow option to preserve everything.

So I hooked everything up and ran it:

mfsbackup -Tao - /dev/hdc | mfsrestore -s 127 -xzpi /dev/hda

except I had it hooked up wwith the above drive letters rreversed hdC being my new drive and hdA being my old drive, which I DON'T THINK should matter right? It was just easier for me to do it this way.

So the command actually looked like this:

mfsbackup -Tao - /dev/hda | mfsrestore -s 127 -xzpi /dev/hdc

It ran OK! everything checked out I got the complete message wiwithow much more recording time I had. I was siked! It was easy.

I shut it down and popped the drive back in, flipped it on, it starting booting and then the GSOD!, it then kept rebooting and repeating.

I started reading posts and soon came across this thread and thought it had to be the swap.

So I started following the instructions for at the top of the thread, except I had no mkswap file on my disk ! (I thought it had EVERYTHING I needed, obviously NOT). I was stumped.

I read another post from a guy who used the above command WITHOUT the -s 127 switch and it worked for him. So I tried it again, and its still doing the same thing.

Some questions:

Am I doing the right thing? using the correct command?
How do I add a swap now? or do I have to do it again with the s -127 switch?
Do I need mkswap and where do I get it?

Some facts:
before doing anything, when the problem first happened, I tried doing a clear and erase everything because I was willing to start over, I don't believe it ever happened because my box would freeze as soon as I tried it. I did see it try come on the screen for a second after my second attempt (without the s -127 switch) but then it just kept rebooting.

At this point, now that I realize I may be able to save everything I want to try, but if I can't I'm not opposed to starting over. How do I do that, if its necessary?

Thanks in advance for any help. My little sons Power Ranger collection is on there and some are rare showings on TV!

[slaponte]

How much swap you need is based on the size of the new drive. I am not sure how big is a DAM10, but you need 1MB for every 2GB of total disk space. (so for a 300GB drive you need 150MB of swap).

After the backup/restore with the proper amount of swap, you need to use TPIP to initialize the swap.

[Blackforge]

My computer has a SATA drive and MFSTools gets stuck while booting. What can I do?

You mean a kernel panic? I was getting this on a couple of machines trying to use weaknees LBA48 and PTVUpgrade LBA48 boot CDs.

This is on a Series 2 240..

First I downloaded a Knoppix DVD (you can use the CD version too)

I then went to:
http://mfstools.sourceforge.net/
Then downloaded the MFSTools statically linked binaries and put them on a floppy.

I also downloaded Tpip and put it on the same floppy
http://www.gratisoft.us/tivo/tpip.html

After you boot into Knoppix

cd /mnt/floppy/mfstools-2.0

./mfstool backup -Tao - /dev/hdX | ./mfstool restore -s 300 -r 4 -xzpi - /dev/hdY

I used a 300MB swap in case I add a second 300GB drive.

Then run tpip v1.2:

cd /mnt/floppy/tpip-1.2

./tpip -1 -s /dev/hdY

I'm doing this because one of my drives in my expanded pair of 120GB died. Doing this from the original drive. 80GB to a 300GB drive.

I based all this on using the following
http://www.gratisoft.us/tivo/bigdisk.html
http://www.dasmonkey.net/lba48.html
http://www.newreleasesvideo.com/hinsdale-how-to/index9.html


I'm in the middle of doing this right now so I'm not sure if it works yet. I don't see why not though...

Don't forget to turn DMA mode on your hard drives/CDs in Knoppix!

Status: Yay! It worked!

[he244]

I recommend the two drive system.

If you back up your base system with season passes, etc. and extract the shows via network that you really want to save occasionally, there is minimal risk or hardship really.

Agree, disagree?

[azitnay]

There's no question that there's minimal risk if you keep a SP list updated and watch or transfer everything you care about, but that doesn't change the fact that two drives means double the hardship. Two drives in effect at least doubles the (admittedly low) hardship, if for no other reason than on average, you'll have to replace and restore the hard drives twice as often.

Given that we can now easily utilize more than 137GB per drive, coupled with the fact that 320GB drives are pretty dirt cheap nowadays, if 320GB is enough space for you, a single drive makes total sense. If you absolutely must have more space than that in a single TiVo, a two-drive system might make sense for cost reasons at the moment, but as larger hard drive prices come down it'll make less and less sense.

That being said, I'm not personally looking to get into a debate on the subject... Everyone has their own opinion, and no one is really wrong.

Drew

[DanDrasin]

Had a power outage, came home, tivo is in death loop hitting the GSOD after a couple seconds on the "just a few more minutes" screen, rebooting incessantly. Googled and found Robert S's post on swap space which sounded like the cause. (I've got a DirectTV/Tive Hughes HDVR2 with a 160GB drive from Weaknees...) So it all kinda made sense.

I followed the directions that Robert S gave and it all went smoothly except that it didn't resolve my problem. (His step E where Tivo is supposed to repair the drive). I was still stuck in the GSOD cycle (let it run for a couple of hours to see if it'd make progress...)

So, i did some more hunting - i mounted the /var partition and read some logs... the tvlog says things like:

String program fsfix
cache at 0x0x5f77dac8, array of 100 entries at 0x0x5f77db50
0.371 seconds: TOTAL for fsfix
fsfix returned non-zero exit code 0x1
fxfix failed to report errors, rebooting...


tverr has endless copies of:

Unable to initialize MFS
Non-deamon Sync() by 128
volume marked as needing database cleanup

So that sounds a lot like the problem that Robert S suggested it was. So of course, the question is "what did i do wrong?" That's probably impossible to answer... but the next question is, can i run fsfix by hand with the drive attached to my linux system?

Dan

[Arcady]

Two hours may not be enough time to let the GSOD fix the drive. Let it run over night.

[DanDrasin]

Two hours may not be enough time to let the GSOD fix the drive. Let it run over night.

It's been running ever since, no progress.
Same stuff showing up in the logs (nothing new...)
Also, contacted Weaknees - they say that the drive that they sent me (which is what was installed) was already at 127M swap size. (and actually jibes with what i think i see in the partition table - though i'm not positive...)

What about running a fsck or some other disk checker. is there a way to rewrite just the partitions that are the OS & app are (not the data) - the answer to this i'm guessing is yes based on how i think it all works, but i haven't been able to piece it together from the how-tos.

(I realize that the likelyhood is that the drive is bad, but before i replace it, i'd like to see if i can recover the data. also, before i invest in another drive, i want to rule out things like the modem, power supply, etc. that seem to also be able to cause this kind of issue (at least some pages say that...))

Any help/advice appreciated.

Dan

[slaponte]

I don't think you can use FSCK. The file system check for Tivo is custom. This goes back to the swap thing, if it doesn't have enough swap, it will never end. Is a particularity of the Tivo fsck... Just an opinion. Sorry, don't have hard data on this.

[renatom]

I am having the same problem. My tivo reboots after the "just a few minutes more.." message. I am upgrading a tivo from 80gb to seagate 400gb.

It seems a common problem and somehow others got it fixed.

Sergio, did you change the -s parameter to create your 340hrs?

this is what I used to create the new disk.
mfsbackup -Tao - /dev/hda | mfsrestore -s 127 -r 4 -xzpi - /dev/hdb

another question from a Newbie: When I run again the same command does it erases my target HD or I have to clean it first?

Thanks in advance

[renatom]

I forgot to mention: I have 7.3 version...

[luan773]

I DL the free MFS Tool 2.0 from Hinsdale, and burned it to CD. All good. But the very 1st command mkdir /mnt/dos gives me an error "could not find kernel image mkdir" . Why? Someone can explain, pls. Thanks. Btw, my Tivo is Series 2 TCD540040. Reason for me to upgrade harddrive is sometimes my TIVO produce bad sound, it's like when you have scratches on CD or tape, very terrible sound. I suspect HD bad and need to replace it, and I would replace it with a Maxtor 250GB. The old Tivo HD is also Maxtor.

2nd, what if I get a new TCD540080 Tivo and swap the HD, does it keep all the config. from new Tivo to my old Tivo? Is it possible? And what if I wanna put the new 250GB HD on top of the pre-config Tivo HD, to expand space, what I have to do? I'm not Linux geek, so someone can give me a good instruction pls. Thanks very much.

[Soloist64]

I have a Series 2 Tivo TCD24004a The hard drive is completely dead. I have a brand new 250 GIG HD Never Used. What are the steps that I need to follow to install the operating system for tivo on this new drive. I don't think I can access the old drive and frankly, I don't want to. Any help would be appreciated.

[kewashi]

I have a Series 2 Tivo TCD24004a The hard drive is completely dead. I have a brand new 250 GIG HD Never Used. What are the steps that I need to follow to install the operating system for tivo on this new drive. I don't think I can access the old drive and frankly, I don't want to. Any help would be appreciated.

Since your old HD is dead you'll have to get a new image. Your best bet is to grab a fresh image from PTV at:

http://downloads.ptvupgrade.com/Merchant2/merchant.mvc?Screen=PROD&Product_Code=ICAKE-S2DT-62

For $20 you can download a fresh image and burn it to your new hard disk.

[wmldwilly]

Wow. This is way more difficult than I thought it would be. I'm stuck in the reboot-loop phase of upping the stock drive in an HR10-250 to a 500gb drive. I've used the hindsale guide and mfstools and all it's companions to get as far as:

1) prepare bootable cd.
2) pull drive, add to pc in ide positions reccomened.
3) create small backup on fat32 disk
4) restore small backup to new 500g drive
5) forget to run/not know I needed to do the tpip thing (like 98% of upgraders it seems...)
6) put the drive back in the HR10-250 w/ jumpers back to master,
7) observe reboot loop.
8) put drive back in PC, leaving it set for master after removing fat32 disk used earlier
9) run tpip.

Here's where I'm totally crapped out:

I do:
#tpip -1 -s /dev/hda

and it replies:
#initializing swap partition (version 0) @ <very long number>, Size = 127 MiB

Put it back in the tivo, still boot-loops.

I'm a little confused at why the -1 option still yields the (version 0) message. Looking closer at tpip's useage it mentions --mkswap and --swaptype={0,1}. Upon trying:

#tpip --mkswap --swaptype=1 /dev/hda

I get:
the same message as before, but with (Version 1) in there.

Put it back in the tivo, and still boot-looping.

So my triplet of questions follows - and PLEASE take pity on me - I've been gearing up for this and reading up on it for a week, and have been working on this from 8pm tonite to 5:45AM now when i'm writing this...it's kicking my arse.

q1) what command do I use to once and for all confirm or deny that the swap partition is there at all? I'm getting the impression that tpip is writing these headers to something mythical.

q2) If the swap partition isn't actually present on the drive at all, how do I actually MAKE one? I'm guessing that's going to be ugly and dangerous, and searches have led me thru threads about copying a swap partition from the CD, but it never quite sounds like what will work for me and an HR10-250.

a3) being that this is an hr10-250 running 3.1.5f, do I have to think about "byteswapping" in this whole procedure? I read in one thread that byteswapping is a series 1 thing, not a series 2 thing. I admit freely I don't completly understand what byteswapping is anyway.

Oh dear gawd please somebody...

Off to pass out,

Willy

Here's where it gets weird

[sync23976]

Just a thought, though it may not matter, have you tried booting the drive in the Tivo with the jumper set to Cable Select (cs)? That is the way mine came from the factory. I am anxious for you to succeed because I am most of the way through to upgrading my HR10-250 to a 500 gig drive and I created a 400 meg swap during the copying. The new drive boots but I have not tried to tpip the swap yet so I'm sure it is functioning with no working swap file, I'm not sure if the stock kernel will recognize such a large swap. I would like to have the larger swap in case it needs to do a mfscheck but maybe it's not worth the hassle.

[ciper]

Even though this thread is ancient I wanted to post something relevant.

My Tivo had two 200gb drives and 127mb of working swap. I stupidly did an "mfsassert -please" not knowing my swap file wasnt large enough. I got stuck in a continuous reboot cycle as fsfix / fsck crashed on every boot.

The interesting thing is that I was able to create an 80mb swapfile on my var partition using only DD for a total of 207mb of swap!
I copied from swap partition 8 and specified to only copy 80 to the file. I then modified rc.sysinit with a "swapon /var/swap" command right after var is mounted (near the beginning). My Tivo accepted this additional swap just fine and the file system check was able to continue.

If I remember right the command I used was
dd if=/dev/hdc8 of=/9/swap bs=1024k count=80
hdc8 being the swap partition of my Tivo's A drive and /9/ being the temporary mount point I used for the var partition (partition 9).

Edit: Although definitely better my Tivo did reboot after staying on the GSOD for 5 hours. Im hoping this is normal and not that I ran out of swap again :(