On Thu, 22 Dec 2005, Connie Sieh wrote:

> Jon,
>
> It was indeed set at 4,  increased it .
>
> Thanks for reporting this issue.

Great.

A few mins ago I switched to try the 
rsync://www.mirrorservice.org/sites/ftp.scientificlinux.org/linux/scientific/ 
mirror (it notionally being closer I thought it might be useful), but 
fetching from it resulted in it deleting a bunch of files (.rpmmacros etc) 
and updating rather a lot of files.  I killed it after 124 of them had 
been changed:

...<snip>...
305/SRPMS/vendor/errata/GFS-6.0.2-12.src.rpm
      1668654 100%    1.08MB/s    0:00:01  (123, 0.9% of 13731)
305/SRPMS/vendor/errata/GFS-6.0.2-24.src.rpm
      1931706 100%    1.76MB/s    0:00:01  (124, 0.9% of 13731)
rsync error: received SIGUSR1 or SIGINT (code 20) at rsync.c(163)

Trying again from rsync://rsync.scientificlinux.org/scientific/ restored 
the missing files and changed the others all back (I can't tell if it 
actually downloaded them all or not)...

...snip...
305/SRPMS/vendor/errata/GFS-6.0.2-12.src.rpm
      1668654 100%    2.02MB/s    0:00:00  (123, 0.9% of 13738)
305/SRPMS/vendor/errata/GFS-6.0.2-24.src.rpm
      1931706 100%    2.13MB/s    0:00:00  (124, 0.9% of 13738)
...

Maybe I should have tried a different mirror!  Does anyone have a simple 
way to test if a mirror contains consistent sets of files?  I suppose I 
could run rsync in test-mode...

  -- Jon