view · edit · print · history

This article pertains to:

  • Unslung 6.x only

I followed the instructions and got upgraded from 5.5 to 6.8. I had a problem with the root account since I had previously changed shells. Apparently the upgrade copies over the old passwd file from /share/flash/conf. I had to re-flash and create a dummy shell program to get in. Everything else worked fine. -Saul (4/14/2006)

Worked like a charm for me. Now, I am trying to get ipkg working. it seems that the link to the web site are broken! Can't get my NFS server running! This is very strange. Can someone fix that link for me?

Upgraded from 5.5 to 6.8 without any problems. The only strange thing I've seen is that the Unslung and the Linksys logo alternates in the web interface. -Jonas Lindstedt (4/17/2006)

Upgraded for 5.5 to 6.8 with out issue after following the dirty upgrade procedure. Changed unlung disk to port2. Just had to change “hdd” to “flash” in the paths of my optional packages’ config files. n8_ball 4/22/06

Upgraded from 5.5 to 6.8. It went smothly until I tried to unsling. I switched my unsling disk to port 2 because I wanted to put a hub on port 1. It appeared drive mounted automatically when I plugged it in after logging in to unsling but it mounted it in read only mode and unsling resulted in error. I tried to repeat process but no matter what I did it mounted it in read only mode. Eventually to get it to unsling I unmounted it and remounted it and then unsling. Then I changed from “hdd” to “flash” in the paths of my optional packages’ config files. Then I had to change rsync for backup to reverse order of drives. Then because I set to back up my next problem was that I couldn't mount disk 2 filesystem because ui thought I was backing up to disk 2 where in fact I was backing up to disk 1. To work around this I created a diversion script to alter smb.conf file also changing hdd to flash. The last problem is backing up to sda1 is not possible since it's not mounted. I'm going to have to mount it somewhere so I can back up to it. garpinc 4/23/06

Upgraded from 5.5 to 6.8. Also had problem with readonly mount describe by previous entry. I determined that quotacheck was the process preventing a umount. I killed that process and then was able to remount rw and then unsling. mda 2006-06-02

Had a couple problems. I caused the first when I moved '/bin' instead of '/share/hdd/data/bin' in step 8. If you do this, please realize it before rebooting as erasing the /bin directory from the nslu2 flash memory causes several problems not the least of which is preventing the slug from rebooting. Argh! Was able to RedBoot and copy 6.8 over and install it. Rebooted, plugged disk into port2 and unslung it. Upon rebooting, the slug came up fine. Enabled telnet, but whenever I telneted in, I would get the motd 'THIS SYSTEM IS CURRENTLY UNSLUNG', then 'connection closed by host' and my connection would be terminated. Powered down, moved the usb disk to my local workstation, copied the backup of my /opt directory over the newly unslung /opt directory so that the bash shell could be loaded, and this solved the problem. Copied my old smb.conf file over, changed the references from 'hdd' to 'flash' and all is good.

I was on 5.5 with a 1gb sd card in a reader in usb1. I mainly run a web server and download torrents using py-bittorent (silently). After reading the instructions and following them closely, I decided to go with dirty upgrade but wish I had tried a straight upgrade... my sd card is now in usb2 I have had to reinstall the packages. One thing the py-bittorent package is missing hence wondering if a straight uprgade would have been better. Glenn (4th June 06)

A disaster. I followed the instructions exactly, however, when I upgraded and plugged in my old drive, I couldn't access the Administration section of the web UI. Neither my password nor the default password worked. This also meant that I couldn't turn on telnet from the web UI. Another person had this problem; see this thread on the Yahoo! group mailing list. Furthermore the public share didn't work -- some kind of missing resource or resource not found error. In the end I was able to recover my data only by using a spare disk in USB Port 2, formatting and unslinging to it, and plugging my old disk into USB Port 1 -- this forced the server to ignore the old disk. I could then mount it and pull files off it. A couple of the larger files were unreadable, which is not surprising, but unreadable in a strange way -- I could access them directly and read their contents (some were disk images), but I could not copy them.

If I had it to do again I'd do a clean install. Slower and more cumbersome, yes, but in my case at least, it would have saved me a day of work (and possibly preserved some data).

Alan (August 2006)

Took a while before figured out how to REALLY do it (because I wouldn't want to loose 150Gb of data ...). Then: 1) took deep breath, 2) took the time to take every single step as described and 3) found out it worked perfectly!

Like Alan, I had a hard time with authentication. Why are there soooo many passwd files? Would someone please write about them on the wiki? Eventually I regained access by booting without the disk attached and using the default password (took me a few goes to get it right).

Aside from this, the rest of the process was quite straightforward. [RickB? Feb'07]

PS - after a couple of tests, it seems /share/hdd/conf/etc is copied to /etc on bootup. So changes to /etc/passwd get replaced. When setting passwords, it's necessary to cp -f /etc/passwd /share/hdd/conf afterwards in order for the changes to persist over reboots. Please can someone confirm this.

view · edit · print · history · Last edited by rickb.
Based on work by rickb, abe, Alan, Glenn, mda, Chad, garpinc, n8_ball, Jonas Lindstedt, titioft, and Saul.
Originally by mwester.
Page last modified on February 05, 2007, at 10:49 PM