Re: [rdiff-backup-users] Rdiff-backup Starting to segfault out of nowhere

[Robert Yoon]

Now I am getting this error. I am assuming in my last email that its 
crapping out somewhere after the /usr/lib/libcrypto.so.0 part.  Now I 
see this message.
What do you guys think?

#rdiff-backup
Traceback (most recent call last):
 File "/usr/bin/rdiff-backup", line 21, in ?
   import rdiff_backup.Main
 File "/usr/local/lib/python2.2/site-packages/rdiff_backup/Main.py", 
line 25, in ?
   import Globals, Time, SetConnections, selection, robust, rpath, \
 File 
"/usr/local/lib/python2.2/site-packages/rdiff_backup/SetConnections.py", 
line 30, in ?
   import Globals, FilenameMapping, connection, rpath
 File 
"/usr/local/lib/python2.2/site-packages/rdiff_backup/FilenameMapping.py", 
line 33, in ?
   import Globals, log, rpath
 File "/usr/local/lib/python2.2/site-packages/rdiff_backup/rpath.py", 
line 38, in ?
   import os, stat, re, sys, shutil, gzip, socket, time
 File "/usr/local/lib/python2.2/socket.py", line 41, in ?
   from _socket import *
ImportError: /usr/lib/libssl.so.0: undefined symbol: sk_new_null



David Kempe wrote:

> Robert Yoon wrote:
>
> > I just recently ran into some problems with one of the hosts i am 
> > using rdiff-backup on. its on redhat 8 and it was just working fine 
> > until recently
> >
> > #rdiff-backup  -V
> > Segmentation fault
>
>
> thats not rdiff-backup's fault i don't think
>
> > --- SIGSEGV (Segmentation fault) ---
> > +++ killed by SIGSEGV +++
>
> did you update the kernel recently? I have seen a similiar crash with 
> a kernel that didn't include NPTL support on a system that used nptl.
> solution is funnily enough to mv /lib/tls /lib/tls.old
> but it could be a hardware problem or anything else in between
>
> dave