1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

MFS Tools 3.2

Discussion in 'TiVo Upgrade Center' started by jkozee, Jun 4, 2015.

  1. Sep 3, 2018 #1101 of 1116
    mr.cello

    mr.cello Member

    89
    0
    Jul 28, 2004
    WA
    Success! (at least so far). The kickstart 58 worked (on the drive/image where I tried the `mfsck -m4`), I have had two successful service connections - will keep an eye on it just to make sure it continues beyond a handful (which was the best I had previously had) and update if things cease to work. Thanks @jmbach for the pointers.

    I am curious however: is kickstart 58 equivalent to running `mfsck -m1` on the drive?
     
    Last edited: Sep 3, 2018
  2. Sep 3, 2018 #1102 of 1116
    jmbach

    jmbach der Neuerer

    2,204
    133
    Jan 1, 2009
    I think what happens is that if there is an issue with the original image, when it rewrites the image to copy and expand, it tries it's best to handle it but doesn't. What that issue is, I do not know. The error handling of the MFS by the TiVo and by MFSTools is definitely not robust.
     
    Last edited: Sep 5, 2018
  3. Sep 3, 2018 #1103 of 1116
    ThAbtO

    ThAbtO TiVoholic by the bay

    9,373
    443
    Apr 6, 2000
    SF Bay Area
    Yes, but if the Tivo software is at least v20.7.2, it would not matter.
    Red Pro line are 7200 RPM drives which are not recommended, more heat, power that the power supply may not handle.
    The regular Red line of drives is what is recommended, 5400 RPM, less power needed.
     
  4. Sep 9, 2018 #1104 of 1116
    mr.cello

    mr.cello Member

    89
    0
    Jul 28, 2004
    WA
    Well about a week in and all appears well. So thanks!
     
  5. Oct 9, 2018 #1105 of 1116
    jerrythea

    jerrythea New Member

    2
    0
    Oct 5, 2018
    Another data point:

    I have confirmed that the provided images, either the VMDK, or the ISO, do not work with real drives, on a physical system connected via a USB3 port, or with USB drives connected in a virtual machine using virtual USB3 (xhci).

    Would get constant I/O errors very shortly after start, and then failure after about 30 minutes.

    I ended up using USB2 on a physical system, and everything worked ok. I did just have to do a KS58 to get the device to properly connect to TiVo online and maintain the product lifetime subscription.

    I was upgrading a 500 GB drive that was about 25% full, to a 3 TB drive.

    It took over 12 hours, which was a real bad surprise.

    Has anyone looked into recompiling the MFSTools source on a more modern Linux kernel and then building a new ISO/VMDK?

    I am sure that this 12 hour time could be cut way down.

    Alternatively, does anyone have a data point using the drives physically connected, internally, via SATA 3?
     
  6. Oct 9, 2018 #1106 of 1116
    dougdingle

    dougdingle HD Tech

    814
    35
    Jul 4, 2007
    West Coast
    It's the USB2 connection that killed you for 12 hours.

    Using internal SATA3, I have moved from a 3TB to a 4TB drive with mfstools in about 8-9 hours. Significantly faster.

    I also rotate two 4TB drives in/out of my Roamio every 60-90 days, and use a standalone cloning dock to make the clones, and that takes roughly 7 hours using WD Red drives.
     
  7. jmbach

    jmbach der Neuerer

    2,204
    133
    Jan 1, 2009
    I have no issue using either images on my computer in a virtual machine.

    I have two laptops I use when running MFSTools. One does not give me any issues with USB3 and the other does. They have different USB chipsets.

    Also it seems to be related to UAS rather than USB3 itself.
     
  8. Nov 6, 2018 #1108 of 1116
    Chards

    Chards New Member

    6
    0
    Jul 31, 2007
    Colorado
    I have a Bolt+ with a dying hard drive. It tripped SMART, having exceeded its reallocated sectors threshold and the Bolt+ flashes all its lights at me. TiVo is going to send a replacement (for $80, not too bad) but I'm attempting to retrieve my programs. I pulled the dying 3TB drive and ran it through two passes of ddrescue on an Ubuntu box to a 4TB drive (what I had around). I only saw read errors/bad blocks at the very end of that process. I grabbed the MFS Tools ISO (from Oct 2017) and ran mfsinfo info on both the dying drive and the 4 TB drive and the output was nearly identical (just some sda vs sdb in a few places). Now I'm at a decision point.

    Option 1. Just plug the 4 TB drive into the Bolt+, go through the kickstarts recommended, and hope it is recovered enough to boot properly with my programs. Recover these programs with either kmttg to a computer or get TiVo to cross-ship me the new box and do a box-to-box transfer. This, I understand, will render the dying drive "dead" to the TiVo as far as program data is concerned.

    Option 2. Zero format the dying drive and hope the reallocated sector count drops below threshold (I've read a few instances where this happened). If it can clear a SMART test after formatting, use mfscopy from the recovery 4 TB drive to the reformatted 3 TB drive and put that in my current Bolt. Recover programs with kmttg or box-to-box. If this option fails for any reason, option 1 is still available.

    Thoughts on these options are appreciated. If I chase option 2, is the ISO from Oct 2017 the "beta" talked about for the Bolt/Bolt+ or do I need something else?
     
  9. Nov 6, 2018 #1109 of 1116
    jmbach

    jmbach der Neuerer

    2,204
    133
    Jan 1, 2009
    I would plug in the 4TB drive and see if it boots up. If you need to do a kickstart, do a KS 58. Once you know everything is functional then find another 3TB drive and dd it back to the new drive. I would not trust that original drive again.
     
  10. Nov 6, 2018 #1110 of 1116
    Chards

    Chards New Member

    6
    0
    Jul 31, 2007
    Colorado
    Thanks, I'll likely do that. TiVo is going to replace the unit for $80, which is probably cheaper and easier than finding a 3TB drive and going through the dd or mfscopy process.
     
  11. Nov 7, 2018 #1111 of 1116
    Chards

    Chards New Member

    6
    0
    Jul 31, 2007
    Colorado
    Update: I connected the 4TB drive to the Bolt+, which took some effort as the 4TB drive is a 3.5" NAS unit. I needed to source some power from a computer as I wasn't sure the Bolt+ could safely power it. It booted up without any issue and did not require any new pairing with cable service (Comcast). Recordings seem to be OK. I ran a KS 58 as recommended after the initial startup.

    I've transferred a few dozen programs to a computer using kmttg and a few of them ended fairly early into the transfer (PS, not TS) process (10 to 25% or so done). I checked some of the incomplete transfers on the Bolt+ and they seemed intact and complete. This could be a kmttg issue, I'll try to transfer one of these to a Premiere I have. If there are other kickstarts I should try that may clear up the transfer process, please advise.
     

    Attached Files:

    Last edited: Nov 7, 2018
  12. Nov 7, 2018 #1112 of 1116
    jmbach

    jmbach der Neuerer

    2,204
    133
    Jan 1, 2009
    No other kickstart codes that I would try. Mightvtryvthe kmttg thread and see if they have any advise.
     
  13. mattack

    mattack Well-Known Member

    23,114
    296
    Apr 9, 2001
    sunnyvale
    Just to be clear, you're sure there isn't a video glitch at the exact spot of the transfer interruption?

    if playing on the Tivo, you see even a brief video/audio glitch, that will cause a transfer to stop.. (due to a bug on the Tivo)
     
  14. Chards

    Chards New Member

    6
    0
    Jul 31, 2007
    Colorado
    I'm not aware of any glitches, mostly because I haven't checked. I decided to transfer the programs I want to save through TiVo's web interface to a Premiere I have rather than via kmttg. I'm still going to get a backup plan in place to a NAS, either with kmttg's auto transfers or pytivo. If I can get one of these working on a Raspberry Pi to a Synology, that would be ideal.
     
  15. mattack

    mattack Well-Known Member

    23,114
    296
    Apr 9, 2001
    sunnyvale
    You said "I checked some of the incomplete transfers on the Bolt+ and they seemed intact and complete."

    If there's a glitch, that's not intact as far as this is concerned.
     
  16. Chards

    Chards New Member

    6
    0
    Jul 31, 2007
    Colorado
    To be clear, my "checking" involved fast forwarding through the source programs with some combination of triple speed fast forward and interval skip forwards until reaching the end. My use of intact meant that the program was there from beginning to end (not truncated). My review would not necessarily reveal any glitches or loss of content. As you suggest, a glitch could interrupt a kmttg transfer. I've encountered this in the past when I used to do kmttg downloads (and encodes) more regularly. Thanks for the feedback.
     

Share This Page