rpm stopped working, what now

[Alexander Apprich]

Jess,

Jess Anderson wrote:
> While installing updates (which I do by downloading the rpms
> from a mirror at cs.wisc.edu into a directory on the target
> machine, then going down the list doing rpm -Fvh <name>),
> something happened while I was doing glibc*, I guess a bad
> signature (like an idiot, I didn't save the output, which
> consisted of endless repetitions of one line). Lots of ctrl-C's
> had no effect, so I just closed the xterm, perhaps a bad move.
> 
> Now rpm just doesn't do *anything*, as far as I can see: no
> output from rpm -q <pkgname> or any other query, nor from rpm
> --rebuilddb.
> 
> rpm --rebuilddb does write a lock file (/var/lock/rpm/transaction)
> and create a directory /var/lib/rpmbuilddb.<pid>,
> but nothing's in it.
> 
> killall rpm doesn't stop it.
> kill <pid> doesn't, either.
> kill -9 <processnumber> does, but I don't like doing that.
> 
> What to do now? Do I need to reinstall rpm itself, rebuild
> the db, etc? Is there a way to do that with anaconda using the
> install disk, or from a different FC3 machine that does have a
> working rpm?
> 
> 

make sure rpm proces is killed, and could you pls check if
/var/lib/rpm/__db.00 exist...

apprich at elmstreet apprich $ ll /var/lib/rpm/__db.00*
-rw-r--r--  1 root root   16384 Mar  7 09:42 /var/lib/rpm/__db.001
-rw-r--r--  1 root root 1318912 Mar  7 09:42 /var/lib/rpm/__db.002
-rw-r--r--  1 root root  450560 Mar  7 09:42 /var/lib/rpm/__db.003

if so, delete them (make a backup if you'd like to) and then run

rpm --rebuilddb

Does your problem then still exist?

Alex