|
Post by pharrisire on Nov 25, 2010 10:06:18 GMT -5
Have the instructions you gave for making KNOS1and 2 on a stick changed for KNOS3? In following the previous instructions, I now get this error - even if I copy/paste your line by line commands:
" Can't open make-usb.sh: No such file or directory "
KNOS-32bit# su KNOS-32bit# cd /usr/local/bin KNOS-32bit# sh make-usb.sh make-usb.sh: Can't open make-usb.sh: No such file or directory KNOS-32bit#
|
|
|
Post by Kevin McAleavey on Nov 27, 2010 1:48:12 GMT -5
Have the instructions you gave for making KNOS1and 2 on a stick changed for KNOS3? In following the previous instructions, I now get this error - even if I copy/paste your line by line commands: " Can't open make-usb.sh: No such file or directory " KNOS-32bit# su KNOS-32bit# cd /usr/local/bin KNOS-32bit# sh make-usb.sh make-usb.sh: Can't open make-usb.sh: No such file or directory KNOS-32bit# Yes ... we did that deliberately since the provisions for making a bootable KNOS USB stick have changed significantly in this release and our previous experiment with allowing this ended up wasting an entire stick for only 2 gigs worth of KNOS. We wanted to use whatever extra space remained on those sticks to allow "swap space" for KNOS to allow it to take on heavier usage before the dreaded "forced reboot/out of memory" when being used heavily. As a result, we had to make some major changes to our code to allow that. Given that we want to work with anyone who wants to make a USB stick KNOS to boot as fast as it would on a hard drive, there are security repercussions since that USB stick could be write-able some day by malware, even though it can't affect KNOS itself, its applications, or its core. However, the idea of write-ability still gives me the shakes. The procedure for doing that is still in there, but the commands and how it's done are now VERY different, and there are now TWO gatekeepers though which enabling it must pass as part of our desire to make sure that our "CANNOT be infected" policy remains valid even on a USB stick or eventually a hard disk should a customer want that despite our concerns. So because of our caution about this "feature" which will eventually pass to our future "hard drive in the sky," we want to provide the instructions and additional support needed to do this through email. Drop me a line at support at knosproject.com and I'll send you the 11 steps to a bootable KNOS USB stick. There are unfinished additional features in there which aren't ready for prime time yet as well which you might notice, and that's the reason why I'd prefer to do this non-publicly since it will change once again on our NEXT test release too. And yes ... ANYONE who wants a bootable USB KNOS is welcome to email us for the instructions as well. Apologies for our "cloak and dagger" requirements, but anyone who's known us and our reputation can attest that we're safe to talk to ...
|
|
|
Post by pharrisire on Nov 27, 2010 10:48:08 GMT -5
Very Much want it!!! Request is on its way!
|
|
|
Post by Kevin McAleavey on Nov 27, 2010 16:30:41 GMT -5
Very Much want it!!! Request is on its way! I'll get that to you later tonight once I'm in the office.
|
|
|
Post by pharrisire on Nov 29, 2010 13:58:52 GMT -5
Just finished making the stick, but before I unmount and reboot I copied one of the error messages that came up - but had no visible effect. Maybe they are both in the new category of "if it boots up then ignore all error messages". But just in case, here is the last one, [glow=red,2,300]KNOS-32bit# exit exit Error setting value: Failed to contact configuration server; some possible causes are that you need to enable TCP/IP networking for ORBit, or you have stale NFS locks due to a system crash. See projects.gnome.org/gconf/ for information. (Details - 1: Failed to get connection to session: Did not receive a reply. Possible causes include: the remote application did not send a reply, the message bus security policy blocked the reply, the reply timeout expired, or the network connection was broken.) We are finished now ... congratulations on your new KNOS USB stick! #### NOTE! Windows will think this disk is empty. It is not! #### KNOS-32bit# [/glow]
|
|
|
Post by pharrisire on Nov 29, 2010 14:29:10 GMT -5
No Joy. It would not boot - it said I was using an out of date version and to go get the new version. Which is what I used to make the stick. I promise I did not yank KNOS3 disc out and stick the KNOS2 disc in at any time whatsoever during the make-the-stick process. I'm gonna go drown my sorrows in the 64bit disc (I hope it works)
|
|
|
Post by Kevin McAleavey on Nov 29, 2010 22:35:05 GMT -5
No Joy. It would not boot - it said I was using an out of date version and to go get the new version. Which is what I used to make the stick. I promise I did not yank KNOS3 disc out and stick the KNOS2 disc in at any time whatsoever during the make-the-stick process. I'm gonna go drown my sorrows in the 64bit disc (I hope it works) Heh. Should work as long as it's not from an expired KNOS ... once we're through all the testing, we intend to get you guys the final "forever build" once all of these little natty weirdies are all pinned down into what justifies a "permanent build." Of course with each of these betas, each one has a new kernel to it until we've caught up with all of the possible hardware combos out there. I'd offer you one of our existing "permabuilds" but you still have a few weirdities left to resolve, and we'd like to get those pinned down first ... As to the weird error message you got back there, I'm guessing that the system got tired of waiting and logged you out of that console and that would cause the "you're not logged in anymore" error once the tasks assigned in there had finished. We do like to close windows and doors behind people when they're done and turn off the lights so that nobody can sneak in.
|
|
|
Post by pharrisire on Nov 30, 2010 13:40:38 GMT -5
I am just chock-full of weirdness-es!
|
|
|
Post by Kevin McAleavey on Dec 3, 2010 23:36:24 GMT -5
I am just chock-full of weirdness-es! Mystery solved! The "make-usb" which shipped with our Beta3 has a problem - am about to release a workaround and will send you an email once it's up on the site. You'll need to download a replacement for what's in KNOS now and run that instead. It'll be real nice, fast and painless ...
|
|
|
Post by pharrisire on Dec 4, 2010 9:57:22 GMT -5
It worked a treat!
Thanks!!
|
|
|
Post by Kevin McAleavey on Dec 5, 2010 1:34:50 GMT -5
It worked a treat! Thanks!! Glad that did it! Once again, words cannot express my embarassment that the USB maker didn't work on the distribution for everybody. We tested it heavily, but not quite heavily enough for our "zero defects" standards here. The new kernel code we're testing in the 3 beta is different from the previous versions, and some slack we had come to count on is no more. Thus a little trick we did to "autosize" the swap space no longer worked the way it used to. But for anyone else who wants to make a stick, good to know we've got that one nailed!
|
|
|
Post by pharrisire on Dec 7, 2010 13:20:12 GMT -5
The stick is working great with the laptop, but..... on the desktop it will not even boot up, just goes so far then says it can't go no farther and will reboot in 15 seconds, which it does. The cycle repeats as long as I'm there to push F8, otherwise it'll just go to xp.
|
|
|
Post by Kevin McAleavey on Dec 8, 2010 4:24:25 GMT -5
The stick is working great with the laptop, but..... on the desktop it will not even boot up, just goes so far then says it can't go no farther and will reboot in 15 seconds, which it does. The cycle repeats as long as I'm there to push F8, otherwise it'll just go to xp. Before I respond, just want to say that what follows is a genuine expression of GRATITUDE, and is not in any way meant to be facetious. THANK YOU for having such whimsically unusual machines! Your machines have uncovered some genuinely oddball characteristics that we would have likely been unable to encounter without your assistance. I *mean* that sincerely! Been scratching many parts of my anatomy over this one in hopes of figuring out how to rework our code to get around this. I'm going to guess that the stick you're using is a "Kingston DataTraveler" or some other stick that has (or once had) "password protection" built into its design such as one of those "U2" sticks. The reason I suspect this is that if it's failing to boot and is actually causing a kernel panic is because it has a weird little design in its own boot sector which is identifying it as either a CDROM or a SCSI RAID disk array rather than as a "regular USB" device. This would result in it reporting itself to the kernel as a "pass0" device rather than a "umass0" device and that could be resulting in that "Bronx handshake" we've got here. I want to rework our "make-usb" to deal with it, but need for you to do a little test in your machine's BIOS if my theory as to your stick is correct. In the BIOS settings for that machine there is probably a "USB LEGACY" setting, and if so, try turning it OFF. See if the stick boots then, and let me know. If it makes no difference, let me know about that stick so I can get my hands on one and see if we can cause our kernel to step past the BIOS telling us that it's a SCSI RAID Array and I'll get you a fix. In our first design, we used the entire disk surface regardless of how large it was, and that of course made it easy to plug in any stick successfully. In our current experimental design, we're trying to make use of any space beyond the 2.4 gigs that KNOS needs in order to provide swap space there to make KNOS more stable, and if the stick is larger than 12GB even make some storage space available. And so, by this design, we've made it a bit more complicated in the booting process and hadn't expected this surprise. If all else fails, I can make available a copy of the old "wasteful" formatter but I'd rather see if we can make our new design work around whatever is happening there first ... Let me know whose stick you're using and also if you're brave enough, see if you can change the USB LEGACY setting in BIOS to see if that makes any difference. Once we have an answer, you can put it back the way it was if you want ... shouldn't matter to Windows either way - that "legacy" setting was required back in the WIn98 days for snotty device drivers, we've achieved indoor plumbing since.
|
|
|
Post by pharrisire on Dec 8, 2010 13:03:25 GMT -5
It is indeed a 4G Kingston DataTraveler, but it never had any data encryption on it that I know of. I'll try the BIOS setting next.....
OK, the setting was Legacy like you thought. Will now remove stick from laptop and try it in desktop again.....
|
|
|
Post by pharrisire on Dec 8, 2010 13:48:15 GMT -5
well rats. No joy after disabling usb Legacy mode in BIOS.
But, at least I got a little more of the info before the reboot:
Fatal trap 12 : Page fault while in Kernel mode . . . Panic : page fault cpuid = 0 uptime 11 seconds Cannot dump. Device not defined or unavailable. Automatic reboot in 15 seconds......
|
|