Yes, I did, and I'm very sorry. It's a very, ah, newbie error, and I'm new here, so I picked this forum. More details:

The system:
* blade server (part of an aging production cluster) - i686
* no CD player
* Redhat 2.4.17-0.18smp

my assets:
* I'm still running a root bash on the machine - redirection, file completion etc all still work
* ash.static is present
* /lib/libc.so.6.old - the file is still intact, just incorrectly named
* rpm is static
* other machines in the cluster are still available

what I tried:
* mv, cp, ln, cat
* ftp in, ftp out
* rcp, scp, mount
* rpm -i: no rpms present on machine - and I see no way of retrieving them
* can't telnet, slogin or even console login anymore - not that those would help.

Normally one would reach for installation disks right about now, I think. But in the absence of a CD player, that's not going to do much good. I think the machine is a goner, but I thought I'd try a cry for help before giving up.

Any thoughts?


Jurgen

Mount the device as an auxiliary filesystem (i.e. not root fs) on another machine and move it from there. I don't know this works with blade servers, but with the regular setup it's be a matter of plugging in the drive to another machine.

If you have a statically linked version of some scripting language (e.g. perl) somewhere, you might be able to use that.