Hello, all:
Here's my issue - I am running puppy 4.2.0 from a USB install. The computer with which I use Puppy the most is a Dell Latitude C400. This lappy cannot boot from USB. No problem - I installed Unetbootin on my Windows partition and Grub4Dos loads the right drivers, opening the gates to Puppy Nirvana...at 12 Mbps. That's right, the C400 sports only a single USB 1.1 port. Solution? A pcmcia card with USB 2.0. Unetbootin still works great, Puppy roars to life and I'm good to go. Until I need to shut down.
If I booted from the pokey built-in USB port, the final save sequence works flawlessly (but painfully slowly). If I'm using the pcmcia card, the card seems to get unloaded too early and /mnt/home is no longer visible. I either get no message whatsoever, or a scrolling list of unavailable inodes. The temporary work-around is to manually save right before I shut down. This seems to work fine, but when I run fsck on my pup_save file, it repairs fragmentation and non-contiguous blocks.
First, am I harming the ext2 file system when /mnt/home is removed prematurely? If not, I can live with my current situation. Otherwise, is there a way to delay or prevent the unloading of the pcmcia card before the save sequence initiates?
I'm happy to provide error messages, but as I'm using my USB install on the C400 this very moment, I will have to wait a bit.
Feel free to use this as a jumping off point for a general overview of the shutdown sequence - it may be of interest to the casual observer.
Thanks!
Delaying pcmcia unloading at shutdown
hd install?
Why not just do a frugal install on your hard drive?
thanks for the question, mulrah
I do have a frugal install of 4.2.1 on my hd, but the USB install allows me to take my OS anywhere. Wherever I find myself, as long as there's a computer that boots from USB, I have my most-often-used applications and settings bundled up in one Puppy.
I should add that when I boot from my HD frugal install, Puppy does not see the pup_save.2fs file on the USB stick (at least when the USB stick is plugged into the pcmcia adapter at boot-time - I don't recall if Puppy can see it when plugged into the on-board USB port).
I do have a frugal install of 4.2.1 on my hd, but the USB install allows me to take my OS anywhere. Wherever I find myself, as long as there's a computer that boots from USB, I have my most-often-used applications and settings bundled up in one Puppy.
I should add that when I boot from my HD frugal install, Puppy does not see the pup_save.2fs file on the USB stick (at least when the USB stick is plugged into the pcmcia adapter at boot-time - I don't recall if Puppy can see it when plugged into the on-board USB port).
hi puppycat - I'm looking for answers as I'm doing the same as you on an IBM T22 - no usb booting, one 1.1 port. I have an hd install of pup420 so have added a menu.lst entry to boot puppy on a stick in a usb2 pcmcia card. (I have a noisy hd I can't be bothered replacing, usb gives me a quiet computer.)
I have tried 3.01, 420, Barry's new 218 and ttuxx 214EX, and they all shut down the pccard early as you describe. For me, the pup_save files all became corrupt and started giving errors.
However, puppy 217.1 works perfectly - for some reason I'm not tech enough to fathom, the card stays mounted until after the session is properly saved. Can anyone explain why? And would it be possible to get the same thing to happen with later pups?
I have tried 3.01, 420, Barry's new 218 and ttuxx 214EX, and they all shut down the pccard early as you describe. For me, the pup_save files all became corrupt and started giving errors.
However, puppy 217.1 works perfectly - for some reason I'm not tech enough to fathom, the card stays mounted until after the session is properly saved. Can anyone explain why? And would it be possible to get the same thing to happen with later pups?
Found a solution!!
Commented out lines 483-488 in /etc/rc.d/rc.shutdown
I have not had a problem since. I believe the service script patch was calling /etc/init.d/rc.pcmcia which unloaded the card before we wanted it to. Please post here if this fix worked for you or if you have a better explanation for why the fix worked (or if you think I'm damaging something by skipping these lines).
Commented out lines 483-488 in /etc/rc.d/rc.shutdown
I have not had a problem since. I believe the service script patch was calling /etc/init.d/rc.pcmcia which unloaded the card before we wanted it to. Please post here if this fix worked for you or if you have a better explanation for why the fix worked (or if you think I'm damaging something by skipping these lines).
working on 420
Hi puppycat. Deleted those lines in shutdown file and my puppy 420 is now saving its sessions properly. Thanks for that one.