On 11/19/06, David G. Miller <dave@xxxxxxxxxxxxx> wrote:
At 4:28 PM +0100 11/19/06, Andras Simon wrote:
> On a fully updated FC5 (as of say, 2 days ago), when started as root,
> rpm hangs even for simple queries. The last few things it does
> (according to strace) are:
>
> open("/var/lib/rpm/Name", O_RDONLY|O_LARGEFILE) = 4
> fcntl64(4, F_SETFD, FD_CLOEXEC) = 0
> fstat64(4, {st_mode=S_IFREG|0644, st_size=40960, ...}) = 0
> futex(0xb7d32900, FUTEX_WAIT, 2, NULL
>
> This doesn't happen if I start rpm as a normal user (and has never
> happened before for root). I see no selinux related messages in the
> logs.
>
> Any ideas?
>
There was an *old* RPM bug (RH9?) that was caused by a corrupted database
that caused rpm to hang. You can try the verifydb and, if necessary, the
rebuilddb. On the other hand, the last item reported by strace is a futex
which says rpm is waiting on something. Check to make sure that there isn't
another copy of rpm running (actually also probably hung and waiting on
something else).
Good idea! There *was* actually a copy of rpm running (looks like it
was started by a cron job last night). It's hard to say whether this
was the culprit, but maybe not, because rpm works now with the rebuilt
db. My guess is that it hung for the same reason, i.e. db corruption.
Thanks,
Andras
