Re: [rdiff-backup-users] Errno 34 on CentOS 5.1

[Alexander 'Leo' Bergolth]

On 11/11/2008 10:06 PM, Andrew Ferguson wrote:
> On Nov 11, 2008, at 3:38 PM, Mike wrote:
>> I recently did a yum update to rdiff-backup-1.2.2-1.el5.rf
>> Since then the following command fails:
>>
>> # rdiff-backup -v9 /etc/ /tmp/etc-test
>>
>> Any ideas?  The details are below.
>>
>> -- Thanks, Mike
>>
>> ...
>> Tue Nov 11 13:33:54 2008  Processing changed file alternatives/antlr
>> Tue Nov 11 13:33:54 2008  Making directory path
>> /tmp/etc-test/rdiff-backup-data/increments/alternatives
>> Tue Nov 11 13:33:54 2008  Regular copying ('alternatives', 'antlr') to
>> /tmp/etc-test/alternatives/rdiff-backup.tmp.98
>> Tue Nov 11 13:33:54 2008  Exception '[Errno 34] Numerical result out
>> of range' raised of class 'exceptions.IOError':
>>
> 
> Also, there's a good chance this is a bug in pyxattr since this
> particular error code is ERANGE. From the getxattr(2) manpage:
> 
>        If the size of the value buffer is too small to hold the result, 
> errno
>        is set to ERANGE.

I think that's the same error I reported earlier today:
A bug in pyxattr < 0.2.2 causes listxattr to report broken values when
called on symlinks, like the ones shown in my debugger output:

-------------------- 8< --------------------
(Pdb) rp.path
'/usr/share/man/man1/lpq.1.gz'
(Pdb) rp.issym()
True
(Pdb) rp.conn.xattr.listxattr(rp.path, rp.issym())
('', '', '', '', '\xff\xff\xff\xff', '', '', '', ')', '', '',
'\xdcd\x9f\t\x04\xea\x7f\t\xdcT\xf2\t')
(Pdb) rp.conn.xattr.listxattr(rp.path, False)
()
-------------------- 8< --------------------

The empty strings above cause the '[Errno 34] Numerical result out of
range' errors when used with getxattr() later in read_from_rp.

Cheers,
--leo
-- 
e-mail   ::: Leo.Bergolth (at) wu-wien.ac.at
fax      ::: +43-1-31336-906050
location ::: Computer Center | Vienna University of Economics | Austria