Re: feature request/change for 2.6.8

2006-03-13 Thread Master Abi
Wayne, How about per module logging as an option, instead of global. Master Erik Jan Tromp wrote: Wayne, I've been experimenting with the {pre,post}-xfer exec options ever since the original diff showed up in patches/ - fantastic stuff. My feature request/change centers around the new RSYNC

rsync a raw device

2006-03-13 Thread Dan Pritts
Hi all - is there some magic flag i'm missing that will allow me to rsync a raw device? Not the dev entry, but the device itself? I have a filesystem (backuppc data store) that does not respond well to directory traversal due to an extremely high number of hard links (which generate lots of di

long file name (91) error

2006-03-13 Thread Doug Lochart
I have looked through the archives for this error and found some instance of it but no solutions.  I have an idea for a workaround but I would like to get some clarification.1)  Unix has as a limit 256 for the length of a file name.  Does this limit also apply to a fully qualified path? 2)  The pat

Re: Default user when name not defined.

2006-03-13 Thread Dave Mielke
[quoted lines by Wayne Davison on 2006/03/13 at 10:05 -0800] >An ID offset is predictable: you simply specify an offset that elevates >the ID numbers to a value above what you use on your systems for real >users. The only potential problem I can see is if you might have >unexpected unreal IDs in

Re: Default user when name not defined.

2006-03-13 Thread Wayne Davison
On Mon, Mar 13, 2006 at 12:46:51PM -0500, Dave Mielke wrote: > No, that'd still be unsafe. It needs to be predictable so that access > rights aren't being inadvertently granted to those who shouldn't have > them. An ID offset is predictable: you simply specify an offset that elevates the ID number

Re: Default user when name not defined.

2006-03-13 Thread Dave Mielke
[quoted lines by Wayne Davison on 2006/03/13 at 09:30 -0800] >offset for numeric IDs, such as "--id-offset=5000". This would map any >ID < 5000 to ID+5000, and any ID >= 5000 to ID-5000 (assuming that >either --numeric-ids was specified or that no matching name was found). >I think I like the lat

Re: copying loopback folders?

2006-03-13 Thread Wayne Davison
On Mon, Mar 13, 2006 at 11:59:59AM -0500, Drew Weaver wrote: > So what ends up happening is that rsync gets to this file called > TMPDIR which is the "image" of the tmp directory and hangs forever. Can you just add --exclude=TMPDIR ? ..wayne.. -- To unsubscribe or change options: https://lists.s

Re: Default user when name not defined.

2006-03-13 Thread Wayne Davison
On Mon, Mar 13, 2006 at 12:00:05PM -0500, Dave Mielke wrote: > Is there a way to define a default user which is to be used when the > actual user doesn't exist at the receiving end? No, there is not at the present. You're suggesting something like a "--missing-uid=nobody" option. Another idea is

DO NOT REPLY [Bug 3392] fuzzy misbehaving if source is a file

2006-03-13 Thread samba-bugs
https://bugzilla.samba.org/show_bug.cgi?id=3392 --- Comment #9 from [EMAIL PROTECTED] 2006-03-13 11:20 MST --- One way to manualy optimize a transfer for a new directory that is related to an old directory is to first copy the old directory using --link-dest, and then copy the new dir

copying loopback folders?

2006-03-13 Thread Drew Weaver
    This is going to sound somewhat ridiculous, it sounds ridiculous to me even saying it; however I’m running into a somewhat unique issue and I need help resolving it.   I am trying to use systemimager which uses rsync exclusively as a means of making ‘images’ of entire systems.

Default user when name not defined.

2006-03-13 Thread Dave Mielke
Have I missed an option somewhere? Is there a way to define a default user which is to be used when the actual user doesn't exist at the receiving end? Right now the action is to copy the user by number, but this is unsafe in certain environments, and it'd be much better to be able to name a defaul

DO NOT REPLY [Bug 3392] fuzzy misbehaving if source is a file

2006-03-13 Thread samba-bugs
https://bugzilla.samba.org/show_bug.cgi?id=3392 [EMAIL PROTECTED] changed: What|Removed |Added Status|REOPENED|RESOLVED Resolution|

Re: is rsync log file thread safe

2006-03-13 Thread Wayne Davison
On Sun, Mar 12, 2006 at 10:56:25PM -0500, Doug Lochart wrote: > This line does not supply a context of its operation. The line includes the PID of the rsync process. Look at the perl script support/logfilter for an example of how to use the PID to find all the lines related to a particular transf

Re: rsync log file question

2006-03-13 Thread Wayne Davison
On Sun, Mar 12, 2006 at 09:59:52PM -0500, Doug Lochart wrote: > Is there a utility out there that will handle the server/client log files > even if you chaneg the format around? You just need to customize the matching code in the rsyncstats script. It ships with support for two log formats (both w

DO NOT REPLY [Bug 3392] fuzzy misbehaving if source is a file

2006-03-13 Thread samba-bugs
https://bugzilla.samba.org/show_bug.cgi?id=3392 [EMAIL PROTECTED] changed: What|Removed |Added Status|CLOSED |REOPENED Resolution|FIXED

DO NOT REPLY [Bug 3392] fuzzy misbehaving if source is a file

2006-03-13 Thread samba-bugs
https://bugzilla.samba.org/show_bug.cgi?id=3392 [EMAIL PROTECTED] changed: What|Removed |Added Status|REOPENED|RESOLVED Resolution|

DO NOT REPLY [Bug 3392] fuzzy misbehaving if source is a file

2006-03-13 Thread samba-bugs
https://bugzilla.samba.org/show_bug.cgi?id=3392 [EMAIL PROTECTED] changed: What|Removed |Added Status|CLOSED |REOPENED Resolution|FIXED

XATTR and extra meta data sync between nodes

2006-03-13 Thread Rakesh Kumar Bhatia
Hallo, Does Rsync support XATTR sync between nodes? Using setfattr -n user.line -v 100 setfattr -n user.creator -v “Rakesh Kumar Bhatia” will create extended attribute for different files. I used rsync -av to copy the files. extended attributes are missing in the destination p

feature request/change for 2.6.8

2006-03-13 Thread Erik Jan Tromp
Wayne, I've been experimenting with the {pre,post}-xfer exec options ever since the original diff showed up in patches/ - fantastic stuff. My feature request/change centers around the new RSYNC_ARG# variables. Would it be possible to make these available as an array (ie: ${RSYNC_ARG[#]}) rather