et.
I do, but I run two backup sets (one for servers, one for workstations.)
I'm not entirely certain the script still works for amrestore. It used
to, but I've enhanced that section recently so that it recognizes
sessions still on the holding disk, and I haven't tested those
ole box to free it.
(Luckily the system itself was IDE-disk based so I could always
gracefully unmount everything except the scsi holding disk.
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
r to 2.4.2p2
>
>> Yesterday, the daily amcheck run gave me the following error:
>>sh: /tmp/amanda/amcheck.main.16130: No such file or directory
>>amcheck: mail command failed: /usr/bin/Mail -s " AMANDA PROBLEM: FIX
>
Isn't "/usr/bin/mail" usally lowercase?
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
ds incremental dumps. On my
previous workstation I had an 8gig partition that was used mostly
for holding disk and cdrom burner workspace, but I had one directory
there I needed to backup. It was a perl database used by the machine's
website and doing a full-backup of the whole direct
sk, not just partition
device names or mount points. But, the version I'm using doesn't
store dump levels for anything except a full partition or mountpoint.
A subdirectory below the mountpoint can be backed up with dump but the
dump timestamps aren't recorded. Or something like
s hostname and it's being pissy about launching
amandad or something.
Then again, I could be thinking of something else completely, too.
It's been months since I touched that box after I got it working.
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
this disk, a low-priority user workstation,
instead of doing the medium and high-priority file servers?
--
Joi EllisSoftware Engineer
Aravox Technologies [EMAIL PROTECTED], [EMAIL PROTECTED]
No matter what we think of Linux versus FreeBSD, etc., the one thing I
reall
On Thu, 18 Jan 2001, Martin Apel wrote:
>On Wed, 17 Jan 2001, Joi Ellis wrote:
>
>>
>> I amflush these to tapes to send to offsite storage.
>> I've already done two tapes from this batch, I have four
>> left to do.
>
>That's a nice idea, but I have
ns every
>> minute, ifconfig'ing down and up the ethernet interface and re-adding
>> the default route. This is what I have to do on my laptop when running
>> RH Linux to keep the ethernet working.
Well, something must be broken with that machine's drivers. I have Re
e.
Robots are a slightly different story. Amanda comes with a number of scripts
you can adapt for use with your particular robot. As long as you can
get a robot control script going, you're all set.
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
thing left to pack!
I amflush these to tapes to send to offsite storage.
I've already done two tapes from this batch, I have four
left to do.
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
HP COLORADO 8GB streamer?
>Thanks.
>
Perhaps
http://amanda.sourceforge.net/cgi-bin/fom?file=111
There are more HP Colorado models on page:
http://amanda.sourceforge.net/cgi-bin/fom?file=46
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
On Mon, 8 Jan 2001, Alexander von Homeyer wrote:
>Just to be sure once again the question:
>does version 2.4.2 support file exclusion on NT clients?
I didn't think Samba's tar command supported exclude lists?
If Samba doesn't, Amanda can't.
--
Joi Ell
't rewind
them.
That's how all the DAT and 8mm drives I've cleaned have worked.
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
;I got the latest dump from sourceforge about 2 weeks ago.
Did you get the static rpm, or the dynamic one? RH5.2 was a libc5
platform, but the current rpms assume you're using glibc.
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
simply as "The Cederqvist". This
is also very good.
I stuck each of them into 2" 3-ring binders and have many bookmarks in
each. If the CVS Desktop Reference doesn't answer the question, one of
the other two does.
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
k.
It really sucks. My older 8500 never did this. I think it's the drive,
not the driver. I've sent email to Exebyte's support address several
times and have never gotten an answer from them.
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
eneralization that Linux users don't install the
>amanda(-no-dee) service config in xinetd.d would be false. =)
Sigh. So Linux users are coming to be held in the same low regard
as AOL users?
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
Run, don't walk, to your local Best Buy and purchase a replacemet for
/dev/hda. Every time I see these errors, they herald the start of much angst
and file lossage.
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
which also consumes half my ram. Running them both causes my machine to
start swapping heavily, and since I spend most of my time doing Java
development, I launch VMWare only when I need to do excel stuff.)
You could try a VMWare-hosted smbtar backup.
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
e that if windows crashes, windows has a
>bug/flaw/lacking (depending on wheather what made it crash was
>use/unexpected_use/downright_cruel_use) there.
Yes. But then, M$ has never felt that a GPF was anything to avoid. :p
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
ones... at
1075 British Pounds per bottle! (Yikes...)
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
On Sun, 3 Dec 2000, Bernhard R. Erdmann wrote:
>I'm pretty sure Amanda shot the tape drive to death! :-)
Let's all chip in and get John a big bottle of Prozac for the holidays!
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
Red Hat systems, mt is provided by its own RPM. It isn't bundled
with amanda. mt is not in 6.2's taper package, either.
mt-st-0.5b-7.i386.rpm
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
manda's fault ...
>
I just searched through samba.org's mail archive, and there are reports
there of NT blue-screening during a smbtar pull from Unix hosts.
I haven't come across a response to that, someone who has more time may
want to search harder through the archives for an answer.
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
n in the first place. So I'm very
>confused - help!
>
>Also I'm not 100% as to what should be going on with .amandahosts, at
>present I have it in ~amanda, containing 'hostname amanda' and this
>seems OK.
>
>Thanks
>- John
>
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
I have a few remote linux boxes
I'd love to backup to my home tape via amanda. (Yeah, over y 64k link
will suck, but it's better than driving across down with the tape drive
and buying extra scsi controllers...)
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
search
engine and our internal bugzilla. Not a real high-traffic server but
I'd still like to get the bugzilla backed up better than it is now.
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
d the last time I had to pull an image >2gig from a
tape:
cd (someplace with enough disk space)
dd if=/dev/nst0 | split -b2000m
cat x?? | restore ...
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
ssignment to $$host?
If it is, comment it out. That's a bug in amstatus where it assumes
every host name is in fact a name and not an ip address.
I think this is fixed in 2.4.2.
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
have tried forcing compression on
>and off with still the same results.
Nah. Tapetype is a oneshot process, it doesn't affect this drive block
size setting issue you're having. (Mind you, it's possible that
changing the blocktype to what you intend to stick with and rerunning
tapetype once to get a new setting wouldn't be a bad idea. My drive's
capacity changed slightly when I switched from 1k to 23k blocks.)
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
g
partition.
I summed everything (ignoring swap, /tmp, and /dump), and then
subtracted the size of /dump. The remainder was 12959432, which is
still bigger than /dump by quite a bit. Unless your compression is
enabled and getting a lot better than 2/1, your full dumps won't fit.
--
Joi
ke full dumps
first, and all of the holding disk is reserved for degraded
incrementals, there is no space on the disk available for non-degraded
full dumps.
Edit your amanda.conf file to say something like "reserve 10".
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
holding disk?
Find the 'reserve' parameter in your amanda.conf file and set it to a
low number. I have mine set to 30, but I have a 30gig holding disk and
a 4 gig tape unit.
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
ng to read it.
Check your system's syslog or kernel log for tape error messages...
Don't forget you have to 'amlabel' each new tape before Amanda will
write to it.
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
HA-2940U2W
>from adaptec.com and it is indeed one of the BIOS settings. It's on by
>default, and I doubt that I changed it, but I'll have to check.
Would enabling disconnect on the scsi controller be a good thing to do
generically?
I have a similar card on my machine, perhap
self is supposed to be world read/write, even user nobody can
write there under most unixes. If /tmp isn't world writable all kinds
of things break. I wouldn't hold this one against Amanda...
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
ay games with
your fstab to give the backup machine root write rights on the NFS mount
so that incrementals can function.
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
admin1.corp.walid.com: [can not read/write /etc/dumpdates: No such
>file or directory]
/etc/dumpdates is provided as part of the dump package, and rpm
undoubtedly replaced your amanda-writable permissions with root-only
permissions, which are the usual default.
Go chown /etc/dumpdates back to use
em is only backing up files
>which are world readable.
User backup should be added to the same group which owns the disks,
and that group should have read or read-write access to the physical
devices.
(I'm not positive about the read-write, but it definately needs read.)
--
Joi Ellis
[EM
ress, ie the Linux flavor
will, but default Solaris flavors won't.
I've found that simply adding entries to /etc/hosts files was sufficient
to get amanda running. As long as each machine can 'ping othermachine'
and there isn't a conflict with DNS entries, it's fine.
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
SQL guru.
$ is a wildcard character for Oracle, I think.
Replace that chain of pattern with the simpler perl pattern:
".\d\d\d"
"." means any one character, "\d" means any digit.
A safer pattern, the one I use, is "DailySet1\d\d".
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
On Sat, 4 Nov 2000, Edwin Chiu wrote:
>Where can I get 2.4.2?
>How stable is it?
>
>If it's in CVS can someone tell me how to check it out?
There is an entry in the online Amanda FAQ with step-by-step
instructions for getting the current CVS snapshot.
--
Joi Ellis
[EMAIL
ork).
Inetd or a close relative is a standard fixture on unix machines. The
only inetd configuration you must do to support amanda is add three
lines to the inetd.conf and kill HUP the inetd process.
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
entry until the file is closed. You want to be watching the free space
drop in "df -k ." rather than watching "du thefile"
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
list
>thanks ahead of time
>
>dmc
Did you check the online faq? There are a pile of DLT tapetypes
in there. It's likely some of them match your drive's specs, if not
its exact brand/model name.
http://www.amanda.org/cgi-bin/fom?file=93
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
ve case notches, your drive will do it, but you may
experience higher error rates if the media isn't up to the task.
On my drives, once a tape is written with compressesion/density
settings, that tape will forever retain those values until forcibly
erased, usually with an 'mt weof'
fig.h.in'
> :
> :
>
>
>Can anybody tell me the right patch instruction (mean the right options or
>better the exact syntax).
>Thanks a lot
>farnk Rippert
>
>
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
otches.
IE I know there are notces in some types of cartridges to indicate
length and/or maximum write density. And there are special notches on
cleaning tapes.
Perhaps your drive isn't cooperating because it doesn't like the media.
Find a new cartridge you know is compatible with
On Fri, 20 Oct 2000, Dean Pentcheff wrote:
>Joi Ellis wrote:
>> On Tue, 17 Oct 2000, Jonathan D. Proulx wrote:
>...
>> I'm also not sure I'd want the identity of my tape backup server
>> visible with simple DNS queries, either. That makes it a juicy
>>
g on a
tape header, an erase won't do it. I'm certain the dd won't...
(4mm DAT of various vintanges, EXB-8500, EXB-8505).
(Did you mean for the dd to overwrite the whole tape? Usually just
whacking the header is sufficient. If not, add a count=NN to it.)
--
Joi Ellis
[EMAIL PROTECTED], http://www.visi.com/~gyles19/
his way the chunks are already compressed,
amanda knows exactly what size they will be on tape and can plan much
better.
When I play games with my own offsite backup extractor/packer and
amflush, I regularly achieve 99.9% tape utilitization using my
standard tapetype. I h
ugh all my tapes, I won't have to do this any longer
since the drive checks a tape for such settings when it loads it.
I use the dd to make sure the whole header is clean, because otherwise
my tape drive is likely to start shoeshining on the tape's physical
header. I don't know why
53 matches
Mail list logo