you
should attach ktrace? Maybe by writing it to /dev/console or to
syslog?
By the way, it's the launching of sendbackup that seems to be failing,
not the launching of dump.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
in the
code to explain this strange formulation? Also, we'll probably need
to do this everywhere we fork and plumb a new process -- in
pipespawn.c at least.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
that's tricky since this is a daemon process..
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
. Is there a
way to slow down the disk IO so that it doesn't wedge the machine?
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
be disabled in the amgtar application -- see
http://wiki.zmanda.com/man/amgtar.8.html
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
, not in my time with Amanda..
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
?
At this point, it does not select only level-0 dumps -- it will vault
everything written in the latest amdump run.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
are handled, but I won't
have a chance to address the higher-level questions of scheduling,
metadata maintenance, tertiary tape cycles, etc.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
the
old libraries (/opt/sfw/lib) by removing it from various LD_foo_PATH
variables (some of which are specified at compile time, IIRC).
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
into the
holding disk -- to do otherwise would be to write the data to disk one
more time than required, which would be inefficient.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
On Wed, Aug 12, 2009 at 4:18 PM, Dustin J. Mitchelldus...@zmanda.com wrote:
It looks like this was a merge error where one change was merged to
the branch without something it depended on. In this case, the
dependency was a renaming of Amanda::Types to Amanda::Header. I'll
adjust amtapetype
On Thu, Aug 13, 2009 at 3:40 PM, Brian Cuttlerbr...@wadsworth.org wrote:
I'm sure its something minor...but I'm not seeing it.
libDebug.so should require libamanda.so -- is that requirement in
place? and satisfied?
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
On Thu, Aug 13, 2009 at 4:45 PM, Brian Cuttlerbr...@wadsworth.org wrote:
I installed on Cascade using the compilation that I created for/on Curie.
The # make install seemed to go well.
Try running ldd on libDebug.so?
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
libDebug.so is just
looking for libamanda.so, and given the symlinks you pasted earlier,
is finding the old libamanda rather than the new one.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
On Wed, Aug 12, 2009 at 7:13 AM, Jean-Louis
Martineaumartin...@zmanda.com wrote:
There is no new snapshot because there is no commit since 20090805.
Ah, so. I've been off my game!
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
amtapetype in that branch to use Amanda::Types.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
Thanks, Charles -- I added a link:
http://wiki.zmanda.com/index.php/Useful_Amanda_Tools
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
. The idea is to
completely jettison the old Changer API before the next relase.
So, please let me know if the snapshots are working for you!
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
snapshots of the latest and greatest Amanda
code. I should have made that clear :)
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
with the results.
The snapshots are here:
http://www.zmanda.com/community-builds.php
and to be clear: please do *not* run snapshots in production!
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
On Tue, Aug 11, 2009 at 12:56 PM, Gene Heskettgene.hesk...@verizon.net wrote:
I have been following the 2.6.2 snapshots, usually within a day or so, for
several months. You've heard about it when I have problems, which isn't
often.
Thanks, Gene -- this has been a great help!
Dustin
--
Open
!
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
, it is GNU tar
break
;;
*)
echo warning..
;;
esac
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
you want.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
? It wrote ~1.8G of data, then stopped. Did
that filesystem run out of space?
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
On Fri, Jul 31, 2009 at 9:11 AM, Gene Heskettgene.hesk...@verizon.net wrote:
Your turn. :)
Can you send me the whole taper log with all of the repeats in it? or
at least the first few MB of it? I suspect there are a few bugs
conspiring here..
Dustin
--
Open Source Storage Engineer
http
forever. I'd like to see those logs
all the same.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
Ah, I see what happened -- the tape doesn't get overwritten, so the
taperscan assumes that it is still useable and feeds it back to the
taper. Rinse, wash, repeat.
I'll get this fixed up, then.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
On Fri, Jul 31, 2009 at 1:09 PM, Gene Heskettgene.hesk...@verizon.net wrote:
Should that be something that an amcheck /config/ could have caught?
No, not necessarily -- note the skipping tape-writeable test (which
had the side-effect of erasing the tape)
Dustin
--
Open Source Storage Engineer
). If you're doing a one-shot
project like this, then Amanda will just be in your way. Instead,
quiesce the server, tar it up to tape or a NAS or what have you, and
then untar it on the new hardware.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
matters but I run most everything on
ext3 filesystems with the relatime option (or noatime
if older).
I've never tested any numbers on which filesystem best suits a VFS
device. Personally, I use ext3, but I would certainly be interested
in some empirical analysis of this topic.
Dustin
--
Open
above?
If you must use a BSD* authentication, make it BSDTCP.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
creative in designing unit tests for the application.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
is currently empty!
I don't know from AIX, but I'll be happy to answer any code-related
questions that might come up while you're working on this.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
, regardless of the reasons.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
?
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
?
If you lost all of your data, you should probably also 'amrmtape' all
of the tapes Amanda still thinks it has.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
On Thu, Jul 16, 2009 at 3:16 PM, Brandon
Metcalfbran...@geronimoalloys.com wrote:
I already emptied out tapelist, so I guess this isn't going to work.
Any manual steps I can take to get going with new level 0 backups?
Blow away all of your old trace logs.
Dustin
--
Open Source Storage
On Wed, Jul 15, 2009 at 10:47 AM, Brian Cuttlerbr...@wadsworth.org wrote:
planner: clep._xerox.1: disk clep:/xerox not in database, skipping it.
Looks like it's a dump stranded in the holding disk?
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
and then warn us that it wasn't in the
disklist ?
Not if it's not in the disklist..
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
On Sat, Jul 11, 2009 at 9:01 PM, Gene Heskettgene.hesk...@verizon.net wrote:
I just grabbed the last snapshot, and the build fails:
Thanks! Dumb typo on my part, a fix for which I just committed.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
in its capabilities.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
away when amandad exits, so it seems that the problem is that
amandad is not exiting when the backup is complete? Or is the backup
not completing?
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
are you using? Can you tell (using truss, maybe?) what
amandad is doing after the backup completes?
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
. The error
reporting from --with-testing was accidentally broken, and we'll get a
fix committed shortly.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
On Wed, Jul 1, 2009 at 10:16 AM, Brian Cuttlerbr...@wadsworth.org wrote:
Does amanda check port restrictions against max_client ? should it ?
This kind of got dropped.. Jean-Louis, what do you think about this
suggestion? Brian, do you want to add a check for this to amcheck?
Dustin
--
Open
. But the zombies should disappear when the parent process
(amandad) exits..
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
like this, I'll be happy to help.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
of Amanda in a
different prefix, and adjust your $PATH accordingly.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
.gz program /usr/bin/xtar
I'm confused -- aside from using xtar and not gtar, what's wrong here?
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
with the case-insensitivity of
the Mac filesystem? What happens if you expand your expressions to
also include lower-case letters?
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
is now an Amanda application, and its interface is
not the same as the tar executable.
I'd be interested in a patch that causes Amanda's amgtar application
to optinonally use 'setar'. I imagine this would be fairly simple?
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
On Mon, Jun 29, 2009 at 9:12 AM, Dustin J. Mitchelldus...@zmanda.com wrote:
It looks like the behavior you want will require a patch. If you make
the permissions a device property, I'll be happy to merge the changes
into Amanda.
Anyone else want to write up this patch? I've got a million
the behavior you want will require a patch. If you make
the permissions a device property, I'll be happy to merge the changes
into Amanda.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
.
This appears in the code:
/* This looks dangerous, but is actually modified by the umask. */
#define VFS_DEVICE_CREAT_MODE 0666
so I think you should look to the process umask to adjust that.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
process for the problem until I did a test run during
the day while they were watching.
Glad that got straightened out.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
handling, see the mtio man page for more details.
You are correct -- amanda no longer needs a non-rewinding tape.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
On Mon, Jun 22, 2009 at 10:26 AM, Steven
Backusbac...@whimsy.med.utah.edu wrote:
I can't find this anywhere in the wiki or other documentaion. Can
you point me in the right direction?
http://wiki.zmanda.com/man/amanda.conf.5.html
look for label_new_tapes
Dustin
--
Open Source Storage
a part of a group that has access to
the devices ('disk' in many cases).
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
in the code/behavior?
Everything -- the Device API was a complete rewrite of all device
handling in Amanda.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
the firewall and still get same error.
I can ping the server with no problem.
See the troubleshooting section of the wiki.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
, but broadcast traffic is
usually counted and reported separately, so I suspect this may,
indeed, be broadcast)
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
and other changer-related stuff in amanda.conf.
Also note that the leading slash is important:
tape:/dev/nst0
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
it
on. The changer web interface also tells me that the drive has
hardware compression off too. Puzzling.
amtapetype detects compression by noting that it can fit more data on
tape when using a fixed byte pattern than when writing pseudo-random
data.
Dustin
--
Open Source Storage Engineer
http
other tape).
Has anyone seen this kind of behavior? I'm planning to build
Amanda::Changer::robot to work with this sort of behavior, but I'd
like to be able to pin it down a bit better first.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
on the client?
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
locating the
server is definitely the simpler solution :)
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
that folks on this mailing list would recommend Amanda..
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
quoting wrong, but that's a different bug).
Here's the tar invocation I learned from Nick Russo lo these many years ago:
tar -cf - $files_to_copy | ssh $remote tar -xf -
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
to arbitrary clouds. If you write an interface from libzcloud to
your cloud system, then Amanda will be able to use it.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
property to the tape device.
I'm not sure how accurately varoius operating systems can determine
the current file, though, so this might be a dangerous property. It's
also currently very difficult to supply properties via amrestore,
since it doesn't have a configuration file.
Dustin
--
Open Source
down the tape to retrieve the image.
Amanda no longer relies on the no-rewind tape device, and always
positions the tape itself.
You may want to consider using a single amrestore invocation to get
the data you need, or using amrecover.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
and things are working.
This seems to be a misfeature of the Solaris services -- it helpfully,
permanently shuts down a service due to a temporary failure, which
seems to be the opposite of resilience. What's the deal?
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
of those issues down to some kind of sqlite stuff
getting confused (which is used by smf(5) IMHO).
This is good info. Robert or Peter, do you want to add it to the
wiki? I feel like we should have a more extensive Amanda on Solaris
section, but I'm certainly unqualified to write it..
Dustin
On Wed, May 20, 2009 at 4:53 PM, Deb Baddorf badd...@fnal.gov wrote:
I've found some reports of GLIB too old errors on initial amanda
compilation,
on Solaris boxes, but I'm having the problem on a Linux box.
What distro? What version of glib is installed? Note that glib != glibc.
Dustin
around a few errors due
to bugs in older versions of glib.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
of variable dumps at the end; *before* those, you will
see the log data. Please examine, and if necessary copy and paste the
last bit of log data into an email.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
On Wed, May 20, 2009 at 7:38 PM, Deb Baddorf badd...@fnal.gov wrote:
conftest.c:253:18: error: glib.h: No such file or directory
This suggests you may not have the glib development package (that
includes the headers) installed. Ah, the joys of binary-only distros
:)
Dustin
--
Open Source
the best way to keep restores quick, and also to keep your
backup size approximately the same from night to night.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
/2cb92fbc5f7f1ad452b9be44670b02d5bd33d194
http://github.com/zmanda/amanda/commit/2cb92fbc5f7f1ad452b9be44670b02d5bd33d194.patch
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
, but it's still a long way off. Beyond the sheer
volume of work to implement it, I think the major challenge will be
handling and storing the metadata describing the location of a
particular tar file without using so much overhead as to render it
unusable.
Dustin
--
Open Source Storage Engineer
http
that oversubscribing your vtapes is a
problematic configuration. If your vtapes are oversubscribed, you
need to pay *very* close attention to what Amanda is doing, and make
adjustments early, before Amanda finds a vtape with 0kb available.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
the whole process for you.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
).
All of this is pretty far off in the future, sadly. I can only code so fast :)
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
-- the relevant Amanda version is 2.5.2p1 for both. So
I would suggest that you look into the changes Ubuntu made between -3
and -4. Perhaps they somehow changed the default for compression?
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
files
-- that's what they're for! If *gnutar* were reading the split files,
there would be a problem, since those files are created with mkstemp
and immediately unlinked.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
if they were interested in writing a
CD-RW Device and the answer was not quite NO, but I doubt there's
been any work done.
This would be a good project for someone with a bit of familiarity with C.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
. This is
done using /dev/nst0. I can manipulate the storageloader using mtx commands
and have been able to write to the drive without any problems.
Is this on Linux? What is the exact error message?
Dustin, do you know if the SAS interface is supported by Amanda?
Anything that works
CDRs, you use 700mb vtapes, and
set runtapes 1.
I'm sure folks looking for truly archival storage know the
implications of their choice of media. A CDR is certainly portable,
and will not be unreadable within the next year or two. Heck, I can
still read 3.5 floppies..
Dustin
--
Open Source
problems.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
. Please post to the
correct forum.
Dustin
Other amanda-users denizens: sorry for the noise -- backupcentral
seems to to provide a way to reply to the sender only.
--
Open Source Storage Engineer
http://www.zmanda.com
the -r flag, and then use dd to strip off the headers and
concatenate the results.
Btw, is there a way to rebuild indexes from tapes?
Amfetchdump has a -i option to do this. I'm not sure it was present in 2.5.1.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
of each file in an archive. There are plans to do this in the
Application API, but much of the machinery has yet to be built.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
be going on?
Hmm, I can only find that in the manpage, not the source. Presumably
it's some old message which has since been removed.
Anyway, probably what it's doing is recovering your file. Amanda
reads the whole dumpfile, even to recover a single file.
Dustin
--
Open Source Storage Engineer
headers on them?
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
have some
context, I may be able to give more detail -- how is gzip being
invoked? Can you use truss or the like to figure out what it's trying
to do?
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
, then you should be calling Maxtor.
Drives go bad, and all of the system-level tools in the world are
basically helpless to prevent it :(
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
a manpage
for it. See amanda-auth(7):
http://wiki.zmanda.com/man/amanda-auth.7.html
Thanks, Paul :)
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
only supported BSD auth, so it's possible you have 2.5.2
configured with a different auth method.
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
On Tue, Apr 14, 2009 at 5:16 AM, Stefan G. Weichinger s...@amanda.org wrote:
Or will they be delayed until the new way of building those pkgs has
been decided (Dustin pointed us to
http://forums.zmanda.com/showthread.php?t=1815 two weeks ago) ?
Checking that thread, it's still a work
601 - 700 of 1254 matches
Mail list logo