Subject: Re: pkg/15501: rsync development branch causes corruption with alpha server
To: Manuel Bouyer <>
From: Greg A. Woods <>
List: netbsd-bugs
Date: 02/26/2002 20:09:30
[ On Tuesday, February 26, 2002 at 19:48:10 (+0100), Manuel Bouyer wrote: ]
> Subject: Re: pkg/15501: rsync development branch causes corruption with alpha server
> On Tue, Feb 26, 2002 at 04:08:59AM -0500, Greg A. Woods wrote:
> > Any idea yet on when might have an updated and
> > un-corrupted copy of the CVS repository again?
> As far as I know anoncvs is ok now.

I don't know anythin about the current stat of anoncvs -- I'm asking
about rsync, which definitely is not OK yet.

> Did  you try 'rsync -c -W' ?

Hmmm.... '-c' isn't going to be very friendly to the server, but I guess
I can try it once:

       -c, --checksum
              This forces the sender to checksum all files  using
              a  128-bit MD4 checksum before transfer. The check-
              sum is then explicitly checked on the receiver  and
              any  files of the same name which already exist and
              have the same checksum and size on the receiver are
              skipped.  This option can be quite slow.

Note that '-W' cannot do anything if there is no need to transfer any
files in the first place:

       -W, --whole-file
              With this option the incremental rsync algorithm is
              not used and the whole file is sent as-is  instead.
              The  transfer  may be faster if this option is used
              when the bandwidth between the  source  and  target
              machines  is  higher  than  the  bandwidth  to disk
              (especially when the "disk" is actually a networked
              file  system).   This  is the default when both the
              source and target are on the local machine.

(and yes, I've confirmed this as much as I can with the rsync source,
though there is some small room for error in my reading of this rather
complex code).

Note also that still fails Alan's test (see his reply
in the PR database) even with '-c':

	$ rsync -v -c -W rsync://,v /tmp/foo,v
	wrote 119 bytes  read 91924 bytes  26298.00 bytes/sec
	total size is 91784  speedup is 1.00
	$ rlog /tmp/foo,v > /dev/null                                     
	rlog: /tmp/foo,v:189: missing 'author' keyword
	rlog aborted

As does too:

	$ rm -f /tmp/foo,v
	$ rsync -v -c -W rsync://,v /tmp/foo,v
	wrote 119 bytes  read 84722 bytes  13052.46 bytes/sec
	total size is 84582  speedup is 1.00
	$ rlog /tmp/foo,v > /dev/null
	rlog: /tmp/foo,v:49: missing date
	rlog aborted

Hmmm.... corrupted, but differently! (well it is a different size too)

> Can you try to rsync from and tell us how it goes ?

First let's try to run Alan's test, with '-c', with that sever too.

	$ rm -f /tmp/foo,v
	$ rsync -v -c -W rsync://,v /tmp/foo,v
	@ERROR: Unknown module 'anoncvs'
	rsync: connection unexpectedly closed (45 bytes read so far)
	rsync error: error in rsync protocol data stream (code 12) at io.c(140)

oops -- maybe not.....

What's the correct incantation for, and why would it be different?
There's no hint that it should be different in the mirrors docs:


I get no luck with finding the CVS repo on rsync://
either.....  From what I can tell from remote it seems there is no copy
of the CVS repo on, even though they claim 58 GB of
NetBSD-related stuff is there.

								Greg A. Woods

+1 416 218-0098;  <>;  <>;  <>
Planix, Inc. <>; VE3TCP; Secrets of the Weird <>