On Monday 26 June 2006 11:12, user100 wrote:
> Hello,
>
> Is anybody running bacula-fd on machines that run Redhat 7.3? I´m afraid
> UTF8 was not the standard all the time on Unix Systems and changing to UTF8
> is not as easy as specifying another "LANG" variable in /etc/sysconfig/i18n
> on older systems. We use a machine with LANG=en_US.iso885915 (RH7.3) that
> carry about ~190 GB data (for the moment) and is not so less used from
> different users (so searching and "hot" renaming all depending filenames
> would be...). There are older systems which may run older Redhat, and AIX
> 4.3 (LANG=en_US) too. If somebody is running bacula on RH 7.3 or AIX 4.3
> and have an idea on howto handle Non-UTF8 data on this system please tell
> me too.
>
> Are there other problems except the "empty folder problem" in wxconsole
> (Restore mode) caused by Non-UTF8 database entries - that would be caused
> by Non-UTF8 data?. I would be happy if at least wxconsole put some message
> out (crap-filenames or a warning) - to avoid stress if something important
> should be restored and it seems the hole folder (!) was not backed up (but
> I hope something is still backed up every time). We use bacula on
> windows-machines too and got NO problems there (even with "Umlauts") - so
> for the moment I think that older Windows (NT4) systems are making less
> problems than older Linux systems...
Try using bconsole for restores. It probably will at least display the
filenames as you see them, though input may be a bit difficult.
By the way, UTF-8 is a superset of US ASCII codes so as long as your filenames
are ASCII you really should not have problems. If you are using a different
code page and non-ASCII characters, then you will have problems if not using
UTF-8.
I am sure you are aware that it is not the quantity of data that is important
but the number of file/directory names that contain non-ASCII characters that
were created under the "old" code page schemes before UTF-8.
Windows has its own set of problems that are equally bad, but at least should
be resolved with later systems and Bacula 1.38.x and later.
A careful examination of the characters in your code page (885915 I never
heard of that) might allow you to make one giant sweep through all your
filenames and convert them to UTF-8 -- however, before doing so, you should
test carefully to ensure that all the software you use handles UTF-8 or
switch to using only ASCII names.
--
Best regards,
Kern
(">
/\
V_V
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
