On Sun, Apr 23, 2006 at 07:15:25PM -0400, Guy Dallaire wrote:
> 2006/4/23, stan <stanb AT panix DOT com>:
> >
> > OK, this starting to get really strange. As I mentioned the machine I
> > normally use for building Solaris 8 (SPARC) binaries has been worked
> > on recently, and I had assumed that my problems were related to this.
>
>
> Who "modified" this machine and what have they done exactly ? If it's an
> outsourcing job, consider flushing them...or talking to someone able to
> flush them, telling your boss that the job they done, rendered your server
> difficult if not impossible to recover
It's a _long_ story. In any case, I rpeated this behavior on a clean
Solaris 8 install tody, so tahe issue becomes moot for thsi discussion.
>
> Any sugestions here? It apears to me that libtool is not doing it's job.
> >
> > BTW, I'm considering building a Soalris boot cd with Amanda on it. Given
> > that,
> > is it possible to build a static version (of the recover toolset)?
>
>
> Hum... I don't think you have to do so, there is something wrong with this
> server for sure, and you'll probably end up having problems with any other
> package you build from source on this machine.
See above.
>
> I've built amanda on solaris 2.6 and 9 without any problem whatsoever.
> Numerous people are probably running it and building it on solaris 7 and 9
> as well.
>
> Someone messed with this server, you got to find what happened and make sure
> it does not happen again.
>
> I recall that the executable were fine after you set back the executable bit
> on lib_gcc, and later that night, it failed. You did not mention if the
> executable bit on libgcc.so was again unset later that night. If so, some
> process is messing with the system.
No, the permissions were fine. The significant difference is that the test
was run from the command line. Also amdnad ran fine when imetd called it, but
it's
child died. Look at the logfile.
--
U.S. Encouraged by Vietnam Vote - Officials Cite 83% Turnout Despite Vietcong
Terror
- New York Times 9/3/1967
|
