Hello,
OK, I now have a valid patch for this, but I am going to hold it a bit longer
before I release it (possibly later today or tomorrow). To properly release
it requires a new version (e.g. 2.0.1) since the problem is in the FD, and
there is no simple patch mechanism for rpms nor for the Win32 FD.
I previously mentioned several workarounds, and by way of this email, I would
like to propose a better set of workarounds:
1. Switch from using "signature=SHA1" to "signature=MD5" these
signatures are not required (except for verifies) and are not currently
used for much.
2. If you wish to keep "signature=SHA1", then follow it immediately
by a "onefs=yes", which is the default and recommended. For example:
FileSet {
Name = "my Set"
Include {
Options {
signature = SHA1
onefs = yes
compression = gzip
}
File = /etc/
File = /var/www
}
}
Should work fine, and is preferable to putting the signature
at the end of the fileset, which might cause the FD to pickup
walk off the end of the options into a bunch of garbage with possible
fatal results.
You might even put the "signature", "onefs" pair at the end of the
Options block, which be even more conservative, but always having
the onefs=yes after the signature=SHA1.
3. If you prefer the patch, please contact me. It does fix the problem
and has passed the (enhanced) regression tests.
Best regards,
Kern
PS: for the purists, I have purposely used the same thread but changed the
subject for clarification, and I have purposely top posted.
On Tuesday 09 January 2007 13:48, Sebastien Guilbaud wrote:
>
> I've partially upgraded to bacula2 (director and sd), all of
> my 1.38 fd's backups report software compression, and all recent
> fd's backups (1.39 and 2.0) report no software compression. I've
> verified the Options { } order trick in Fileset resource, all
> configuration files are almost the same :
>
> FileSet {
> Name = "my Set"
> Include {
> Options {
> signature = SHA1
> compression = gzip
> }
> File = /etc/
> File = /var/www
> }
> }
>
> all fd's are correctly linked against zlib :
>
> linux-gate.so.1 => (0xffffe000)
> libz.so.1 => /usr/lib/libz.so.1 (0x4001d000)
> libpthread.so.0 => /lib/tls/libpthread.so.0 (0x4002f000)
> libwrap.so.0 => /lib/libwrap.so.0 (0x40041000)
> libstdc++.so.6 => /usr/lib/libstdc++.so.6 (0x4004a000)
> libm.so.6 => /lib/tls/libm.so.6 (0x4011a000)
> libgcc_s.so.1 => /lib/libgcc_s.so.1 (0x40140000)
> libc.so.6 => /lib/tls/libc.so.6 (0x4014b000)
> /lib/ld-linux.so.2 (0x40000000)
> libnsl.so.1 => /lib/tls/libnsl.so.1 (0x40283000)
>
> I've tried to run both fds in debug mode, teeing stdout
> to a file, but I found no references to compression
> events, no way to understand why I get
> Software Compression: 30.9 %
> and
> Software Compression: None
>
> any hints ??
>
> have a nice day !
>
> --
> Sébastien Guilbaud
>
> -------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to share your
> opinions on IT & business topics through brief surveys - and earn cash
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
> _______________________________________________
> Bacula-users mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/bacula-users
>
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Bacula-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bacula-users
