Amanda Backup Strategy
I have a tapeless back-up config here with 2.4.3. I have 30 virtual tapes, and so have amanda.conf values of dumpcycle 30 days and tapecycle 30 tapes Now, man amanda tells me that this dumpcycle value should mean that each disk using this set of options will get a full backup at least this often i.e at least every 30 days. However, when I look an amoverview I see that several disks are getting a full backup EVERY OTHER DAY! What's the idea of this? I'll grant it's within the letter of the manual, but it's hardly space-efficient is it? Is there any way I can force amanda to only do the full backup at most n times during the dumpcycle? Also, I see that other disks are getting repeated level 1 incrementals. I would be happier if it rather did level 1 then 2 then 3 etc. and no backup if no files had changed. Can I specify this behaviour somehow? I know it will make a full restore a little messier but I simply don't have unlimited disk space for amanda to use. Thanks in anticipation for advice. Regards, Ralph Bearpark.
Re: Q: simple but how ?
Hi!Toomas,
I've indicated 40Gb because my PowerVault is configured as yours, to compress data by
hardware ( I think in a 2:1 ratio by default). I don't use any software to compress
data, and all the data I should backup are locally to server. Do you think in this
scenario will I have problems with amanda if I indicate 40Gb -tapes in my descriptors
(or seemed, 35Gb if 40Gb is too much optimistic) ?
I will make use of your filemark and speed values, much thanks ;-)
Best regards,
- Original Message -
From: Toomas Aas [EMAIL PROTECTED]
To: Ra.l Wild-Spain [EMAIL PROTECTED]; [EMAIL PROTECTED]
Sent: Tuesday, February 25, 2003 5:56 PM
Subject: Re: Q: simple but how ?
Hi!
define tapetype DDS-4-40{
comment DDS-4 PowerVault with compression
DDS-4-20
length 4 mbytes
}
This length number is probably too optimistic. This assumes that your
data compresses with a 2:1 ratio, which is rarely seen in real life.
FWIW, here's my tapetype with DDS4 drive using hardware compression:
define tapetype DDS4 {
comment DDS4 using hardware compression
length 25000 Mbytes
filemark 0 Kbytes
speed 2816 Kbytes
}
The length number is really just a guess (I think I underestimate the
compression a little, but since I only have 10 GB of data this is not
an issue for me right now). The speed number is taken from avg tp
write rate as reported by Amanda's nightly run.
--
Toomas Aas | [EMAIL PROTECTED] | http://www.raad.tartu.ee/~toomas/
* One way to be happy ever after is not to be after too much.
Qualstar TLS-4480 barcode reader
I'm using Mark Holm's chg-qs-mtx script to start my Backup with Amanda. All works fine, but have problems to use the barcode reader. Is there a way to use the barcode reader for faster library operations ? archiver:/# mtx status Storage Changer /dev/changer:1 Drives, 43 Slots ( 1 Import/Export ) Data Transfer Element 0:Full (Storage Element 27 Loaded):VolumeTag = A00026 Storage Element 1:Full :VolumeTag=A0 Storage Element 2:Full :VolumeTag=A1 Storage Element 3:Full :VolumeTag=A2 Storage Element 4:Full :VolumeTag=A3 Storage Element 5:Full :VolumeTag=A4 . . . . . . The VolumeTag is the barcode readers information. Is there a use for this information with Amanada or mtx to load the right tape Amanada want to use next ? Thanks in advance Daniel
RE: Help getting tape changer to work on Solaris 2.8
Jon Thanks for the reply. I'm embarrassed that I missed so much essential info out of my initial email :( (I work in support, I should have known better) Thought 1, its good to specify which amanda version you are using. build: VERSION=Amanda-2.4.3 Thought 2. you did not say if you had a changer, make, model, /dev entry. If not, you needn't use any changer entries Overland XB with a DLT700 drive. dlt /dev/rmt/1n (I've had the tape respond to this, so i know its right) changer /dev/sq/c3t610 (created using Veritas stuff, I can't vouch for it) Thought 3. you have changerfile set to chg-multi. that would suggest your tpchanger should also be chg-multi Done that, now playing with the chg-multi.conf file Thought 4. for my Solaris system and drive chg-mtx worked fine. others have used chg-zd-mtx (note you will have to obtain and install mtx). Still others have configured the sgen (generic scsi?) driver and used chg-scsi, or the chg-mtx's I explored mtx and sgen, and was unable to get either to create a device for me. Thought 5. /dev/rmt/1n ? Do you have another tape drive at 0? /dev/rmt/1n responds as the tape drive, but I don't have anything I can get to respond at /dev/rmt/0n, even though all the files are there in /dev/rmt/ Is this something I should investigate, or can I ignore them? Thought 6. the device /dev/rmt/1n is choosing the default density and/or compression for your tape drive. If that is what you want, fine, otherwise check into 1ln, 1mn, ... Ah, wondered what they all were! I'll investigate, thanks for the heads up Cheers Nige Here's the current error :- $amtape Daily reset amtape: no tpchanger specified in /usr/local/etc/amanda/Daily/amanda.conf $ Here is the relevant section of the amanda.conf file :- #runtapes 1# number of tapes to be used in a single run of amdump #tpchanger chg-manual# the tape-changer glue script tapedev /dev/rmt/1n# the no-rewind tape device to be used #rawtapedev /dev/null# the raw device to be used (ftape only) changerfile /usr/local/share/amanda/Daily/chg-multi.conf #changerfile /usr/adm/amanda/DailySet1/changer-status #changerfile /usr/local/etc/amanda/DailySet1/changer.conf #changerdev /dev/null Assuming that I need to uncomment the tpchanger line, what should I point it at? Am I right customising and using chg-multi.conf? Any thoughts/guidance appreciated.
Newbie question: merge incremental backup to full one
Hi, How can I merge incremental backup to full backup without using AMANDA ? Thanks in advance. ___ Do You Yahoo!? -- Une adresse @yahoo.fr gratuite et en français ! Yahoo! Mail : http://fr.mail.yahoo.com
Re: Q: simple but how ?
Hi! I've indicated 40Gb because my PowerVault is configured as yours, to compress data by hardware ( I think in a 2:1 ratio by default). Your drive is configured to do hardware compression, period. The data compression rate cannot be configured. The drive just does its best. Drive manufacturers like to advertise that their drives compress data by 2:1 ratio, but that's just what it is - advertising. The compression ratio that you actually get depends of the nature of your data. If the data is all text files and such, you might get even much better compression ratio than 2:1. If, on the other hand, you are backing up mostly .mp3, .jpg and .tar.gz files then you'll get no compression at all, because the data is already compressed. In fact, applying hardware compression to such files might even make the backup bigger than the original. The compression ratio that people who back up entire servers with various types of files - some better compressible, some worse - is typically somewhere around 1.4:1. -- Toomas Aas | [EMAIL PROTECTED] | http://www.raad.tartu.ee/~toomas/ * RUNTIME ERROR 6D at 417A:32CF : Incompetent user
Re: Qualstar TLS-4480 barcode reader
Hello1 We use a TLS-4480, with the chg-zd-mtx that I found someware :-) and it working well (as long as our home printed labels are readable). You can download it at http://persson.cx/code/chg-zd-mtx.gz /jp On Wed, Feb 26, 2003 at 10:27:16AM +0100, Daniel Gruendler wrote: I'm using Mark Holm's chg-qs-mtx script to start my Backup with Amanda. All works fine, but have problems to use the barcode reader. Is there a way to use the barcode reader for faster library operations ? archiver:/# mtx status Storage Changer /dev/changer:1 Drives, 43 Slots ( 1 Import/Export ) Data Transfer Element 0:Full (Storage Element 27 Loaded):VolumeTag = A00026 Storage Element 1:Full :VolumeTag=A0 Storage Element 2:Full :VolumeTag=A1 Storage Element 3:Full :VolumeTag=A2 Storage Element 4:Full :VolumeTag=A3 Storage Element 5:Full :VolumeTag=A4 . . . . . . The VolumeTag is the barcode readers information. Is there a use for this information with Amanada or mtx to load the right tape Amanada want to use next ? Thanks in advance Daniel -- jens persson # In the face of ambiguity, refuse the [EMAIL PROTECTED]# temptation to guess. Mäster Olofsväg 24 # -- Tim Peters 12th Pythonic Thesis S-224 66 LUND;SWEDEN #
Re: results missing
Hi! I guess, the problem kinda solved itself, amanda just didn't like the dumptype always-full Konni
Re: Q: simple but how ?
On Wed February 26 2003 04:40, Raúl Wild-Spain wrote:
Hi! Jon
define tapetype DDS-4-40{
comment DDS-4 PowerVault with compression
Is this hardware compression? If using software compression,
the capacity is still to be listed as ~20GB.
Yes! it is. I think my PowerVault is configured to compress by
hardware in a ratio of 2:1 (by default). I this case do you think
can I make use of these values? are these right?
DDS-4-20
I know dumptypes can include other dumptypes, but can tapetypes
do so also. Besides, the speed would differ for HW compression
on and off. Then again, the speed value is unused, so what's it
matter :) I don't know if filemark size would change.
I saw something like these in my debian-amanda docs ... I will
review it,
Do I suppose this will work properly? Have anyone experience
with DDS-4 tapes and amanda?
Lots of people.
;-))
dumpcycle 0
runspercycle 2
Can you fit two runs into a zero size cycle?
I saw in the documentation that a dumpcycle value of 0 indicates
to amanda run a full backup with every amdump. So I've tried to
indicate to amanda by rotating my 2 tapes you must to do a full
backup every amdump.
I want that amanda warns by mail ... needed tape Backup01
needed tape Backup02 (rotated) and it does a full backup every
amdump with every tape... how to accomplish this?
Set 'tapecycle' to the number of tapes you actually have.
And I concur with the rest of the advice Toomas gave you. You'll
have less trouble in the long run if you shut the hardware
compression off forever, then use software compression only for
those DLE's that will actually compress. You'll get more on the
tape as its better compression, and amanda will know exactly how
much she can put on a tape instead of the SWAG* method you have to
use when using the hardware compressor.
Be aware that turning the compression off on tapes that have already
been written to with it on takes a bit of a walk to do since the
tape, when loaded, tells the drive I'm compressed, and that turns
the drives compression back on regardless of the dip switch
settings.
You basicly have to turn it off when the tape is fully rewound, then
write, using dd, enough data from /dev/zero to force the drive to
flush its buffers, at which point it will update that flag in the
tape header and that tape then becomes an un-compressed tape.
I surround that with some other dd stuff to extract the tape label
first, then do the uncompress routine, rewind and re-write the
label block. That way the stuff in the middle doesn't play with
the old girls mind when she cannot find a tape she knows full well
she did label.
That is going to take a long time to convince your boss that
amanda is working. One run/week? For testing/demo, why not run
it nightly or business nightly?
Amanda works alright, if he has questions, have him join the list
and ask some of us longtime users what we think. We're of course
biased, but the 2 main alternatives are both outragiously priced
compared to free, and considerable more labor intensive on a day to
day basis... Amanda, when properly setup to be run in the hours we
sleep, requires only that the right tapes be in the magazine wheh
she needs them.
Any recovery operation with any software is of course rather labor
intensive, but can be accomplished with nothing more than a bare
metal fresh os install that includes tar, gzip and dd in most
cases. Anything else is just window dressing if push really does
come to shove.
while I haven't the rest of tapes I must to do backups not only
for testing but operative. I thinked that at least doing a
weekly full backup are ok ... at the moment.
Best regards,
SWAG* Scientific Wild Assed Guess, your trivia fact for the day :-)
--
Cheers, Gene
AMD [EMAIL PROTECTED] 320M
[EMAIL PROTECTED] 512M
99.23% setiathome rank, not too shabby for a WV hillbilly
Re: Newbie question: merge incremental backup to full one
On Wed February 26 2003 06:26, Kablan BOGNINI wrote: Hi, How can I merge incremental backup to full backup without using AMANDA ? On the same tape? For dependability reasons, amanda won't do that because she has no way of knowing when next restarted, where on the tape and how much of it has been used. In between sessions, someone may have ejected the tape and re-inserted it, which would rewind it. If amanda then does an incremental on the same tape as your only full, the full is over-written and lost. Not a pleasant scene with the boss if you should need to do a full recovery 3 hours after that incremental was done. Tape is cheap when compared to the cost of lost data and the cost of regenerating that lost data from scratch. Consider if the lost data is your only full, uptodate copy of your accounts receivable... Buy the tapes. -- Cheers, Gene AMD [EMAIL PROTECTED] 320M [EMAIL PROTECTED] 512M 99.23% setiathome rank, not too shabby for a WV hillbilly
Re: Backup and recovery CD
On Tue, Feb 25, 2003 at 05:26:23PM -0500, Matt Hyclak wrote: On Tue, Feb 25, 2003 at 09:32:54PM +0100, Martin Schwarz enlightened us: One of my colleagues has built a bootable rescue CD based on the Knoppix Gnu/Linux distribution (see www.knoppix.net if you don't know Knoppix). His CD includes an amanda client besides our usual tools and goodies. I Any chance we might be able to download said CD anywhere? I've not played with Knoppix, so I don't know how easy it is to add software to the disk... Sorry, it's not available for download atm. But I hear it wasn't that hard to customize the Knoppix CD. It's based on Debian, so the Debian package management tools should work for adding stuff to it and removeing unneeded things. Perhaps amanda client was even already included in the latest version of the original Knoppix distribution, not sure... There are some infos on customizing/remastering Knoppix CDs at http://www.knoppix.net/docs/index.php/KnoppixCustomizations Knoppix Lite ftp://ftp.es.debian.org/pub/miniKnoppix looks like a nice starting point for an amanda rescue CD. Hth, Martin -- [EMAIL PROTECTED] toplink-plannet GmbH Network Operations EngineerSchönfeldstraße 8 Tel +49 [0] 721 6636-0 D-76131 Karlsruhe Fax +49 [0] 721 6636-199 http://www.toplink-plannet.de/
Re: Qualstar TLS-4480 barcode reader
On 26 Feb 2003 at 10:27am, Daniel Gruendler wrote The VolumeTag is the barcode readers information. Is there a use for this information with Amanada or mtx to load the right tape Amanada want to use next ? I haven't sat next to my changer to confirm this, but I'm pretty sure 'amtape $CONFIG label $LABEL` uses the barcode information to load the right tape without trying every one. I'm using an Overland Library Pro AIT3 library with chg-zd-mtx on Linux. -- Joshua Baker-LePain Department of Biomedical Engineering Duke University
Help With Restore
I'm trying to restore a directory on one of our servers. First, I want the directory to reside in a temporary location before I move it to its original location. I ran the following command in the temp location: amrestore /dev/nst0 hostname /.1/shares Now I have a file in my temp directory named hostname._.1_shares.20030206.0. I've tried to tar it to expand the contents but the server just sits there, even when running tar overnight. I used: tar xvpS hostname._.1_shares.20030206.0 What am I doing wrong? I need to be able to pull a certain directory out and trash the rest. Any help will be appreciated.
Re: Amanda Backup Strategy
Linebreaks, man, use some line breaks... Sorry, dude, sorry... Not that I know of, but in the 2.4.4 ReleaseNotes it says that it tries to reduce the number of level 0s per host each night. I guess that *might* help. I could probably get it down to a full backup every week. Would I just need 2.4.4 on the server? Or the clients too? If you want amanda to use less disk space, crank down your 'tapelength'. Since you're using diskless backup, you can define whatever size tapes you want. True, but I have 18GB over 30 tapes. So, to be completely safe from Amanda filling the disk with unnecessary full backups, I'd have to set the tapelength to 600MB, which, unfortunately, wouldn't be enough for any single client full backup. Is there a wishlist for 2.4.5 anywhere? Regards, Ralph.
Re: Amanda Backup Strategy
Linebreaks, man, use some line breaks... On Wed, 26 Feb 2003 at 4:10am, Ralph Bearpark wrote However, when I look an amoverview I see that several disks are getting a full backup EVERY OTHER DAY! What's the idea of this? I'll grant it's within the letter of the manual, but it's hardly space-efficient is it? The goal of amanda is to equalize the amount of data backed up each night while staying within the bounds set by dumpcycle, runspercycle, and tapecycle. If this means it promotes dumps, it does it. Is there any way I can force amanda to only do the full backup at most n times during the dumpcycle? Not that I know of, but in the 2.4.4 ReleaseNotes it says that it tries to reduce the number of level 0s per host each night. Also, I see that other disks are getting repeated level 1 incrementals. I would be happier if it rather did level 1 then 2 then 3 etc. and no backup if no files had changed. Can I specify this behaviour somehow? You can play with bumpsize and bumpmult, but the extra hassle restoring really isn't worth it, IMO. I know it will make a full restore a little messier but I simply don't have unlimited disk space for amanda to use. If you want amanda to use less disk space, crank down your 'tapelength'. Since you're using diskless backup, you can define whatever size tapes you want. -- Joshua Baker-LePain Department of Biomedical Engineering Duke University
Re: Amanda Backup Strategy
On Wed, 26 Feb 2003 at 10:33am, Ralph Bearpark wrote Not that I know of, but in the 2.4.4 ReleaseNotes it says that it tries to reduce the number of level 0s per host each night. I guess that *might* help. I could probably get it down to a full backup every week. Would I just need 2.4.4 on the server? Or the clients too? Just on the server -- that's the job of planner. If you want amanda to use less disk space, crank down your 'tapelength'. Since you're using diskless backup, you can define whatever size tapes you want. True, but I have 18GB over 30 tapes. So, to be completely safe from Amanda filling the disk with unnecessary full backups, I'd have to set the tapelength to 600MB, which, unfortunately, wouldn't be enough for any single client full backup. Single *client* or single *disk*? Remember, you can always split up your filesystems into multiple disklist entries using tar. Then you could reduce your tapelength to have *just enough* to fit everything once. -- Joshua Baker-LePain Department of Biomedical Engineering Duke University
Re: Amanda Backup Strategy
On Wed, 26 Feb 2003 at 11:05am, Ralph Bearpark wrote Remember, you can always split up your filesystems into multiple disklist entries using tar. Then you could reduce your tapelength to have *just enough* to fit everything once. Hmm, dividing my clients into multiple DLEs that I hope will never exceed 600MB ... that's an ugly concept. I assume that Amanda is clever enough to not try to put, say, two 500MB DLE backups on a single 600MB tape? Yep. Amanda gets estimates for all the DLEs at multiple incremental levels, and figures out the best thing to do within the tapelength it has. -- Joshua Baker-LePain Department of Biomedical Engineering Duke University
RE: Amanda Backup Strategy
On Wed, 26 Feb 2003 at 11:16am, Ralph Bearpark wrote Yep. Amanda gets estimates for all the DLEs at multiple incremental levels, and figures out the best thing to do within the tapelength it has. Even to the extent of not doing any sort of backup at all on some or many of the DLEs? If so, how on earth does it decide which? Yep. But, when it does that, you'll get big fat warnings such as: FAILURE AND STRANGE DUMP SUMMARY: $CLIENT $DISK lev 1 FAILED [dumps way too big, must skip incremental dumps] -- Joshua Baker-LePain Department of Biomedical Engineering Duke University
Re: Help With Restore
On 26 Feb 2003 at 9:12am, Kenneth Sullivan wrote Now I have a file in my temp directory named hostname._.1_shares.20030206.0. I've tried to tar it to expand the contents but the server just sits there, even when running tar overnight. I used: tar xvpS hostname._.1_shares.20030206.0 What am I doing wrong? I need to be able to pull a certain directory out and trash the rest. Any help will be appreciated. You need to give tar the 'f' option to tell it what file to read from: tar xvpSf hostname._.1_shares.20030206.0 It was waiting for something to come in from stdin. -- Joshua Baker-LePain Department of Biomedical Engineering Duke University
Re: Newbie question: merge incremental backup to full one
Apologizes for the post because i don't give enough info. Here is. I have make backup on two different tapes say tape A and B from server_one. A is holding the full backup and B the incremental one. I have another server_two not using amanda where i want to restore those data ? What must i do to get all my data(full+incremental) restore on that server ? --- Gene Heskett [EMAIL PROTECTED] a écrit : On Wed February 26 2003 06:26, Kablan BOGNINI wrote: Hi, How can I merge incremental backup to full backup without using AMANDA ? On the same tape? For dependability reasons, amanda won't do that because she has no way of knowing when next restarted, where on the tape and how much of it has been used. In between sessions, someone may have ejected the tape and re-inserted it, which would rewind it. If amanda then does an incremental on the same tape as your only full, the full is over-written and lost. Not a pleasant scene with the boss if you should need to do a full recovery 3 hours after that incremental was done. Tape is cheap when compared to the cost of lost data and the cost of regenerating that lost data from scratch. Consider if the lost data is your only full, uptodate copy of your accounts receivable... Buy the tapes. -- Cheers, Gene AMD [EMAIL PROTECTED] 320M [EMAIL PROTECTED] 512M 99.23% setiathome rank, not too shabby for a WV hillbilly ___ Do You Yahoo!? -- Une adresse @yahoo.fr gratuite et en français ! Yahoo! Mail : http://fr.mail.yahoo.com
Re: Amanda Backup Strategy
Remember, you can always split up your filesystems into multiple disklist entries using tar. Then you could reduce your tapelength to have *just enough* to fit everything once. Hmm, dividing my clients into multiple DLEs that I hope will never exceed 600MB ... that's an ugly concept. I assume that Amanda is clever enough to not try to put, say, two 500MB DLE backups on a single 600MB tape? Regards, Ralph.
RE: Amanda Backup Strategy
Yep. Amanda gets estimates for all the DLEs at multiple incremental levels, and figures out the best thing to do within the tapelength it has. Even to the extent of not doing any sort of backup at all on some or many of the DLEs? If so, how on earth does it decide which? (If this is all documented somewhere, please tell me where and I'll stop asking my stupid questions! :-)) Regards, Ralph.
ERROR: NAK crescent: amandad busy
Does anyone know why amandad would not be going away after being used? I'm trying to back up my server from itself (crescent) and am getting this: ERROR: NAK crescent: amandad busy Also, one other client is giving the same error. The /etc/inetd.conf entry looks like this: amanda dgram udp wait backup /usr/sbin/tcpd /usr/lib/amanda/amandad The server and clients are Debian 3.0, the server software is 2.4.3, but the client is the package version that comes with Debian 3.0, which is 2.4.2p2. Any ideas would be appreciated. -- Terri
Moving to production
Hello all! I read the FAQ about moving from development to production usage. http://amanda.sourceforge.net/cgi-bin/fom?_highlightWords=tapelistfile=34 That answer doesn't mention what to do with the 'index' directory -- do I need to remove this or leave it as it is? Does it make a difference now that I have new tapes? -- David Olbersen iGuard Engineer 11415 West Bernardo Court San Diego, CA 92127 1-858-676-2277 x2152
Re: Newbie question: merge incremental backup to full one
On Wed, 26 Feb 2003 at 5:15pm, Kablan BOGNINI wrote Apologizes for the post because i don't give enough info. Here is. I have make backup on two different tapes say tape A and B from server_one. A is holding the full backup and B the incremental one. I have another server_two not using amanda where i want to restore those data ? What must i do to get all my data(full+incremental) restore on that server ? Extract the full backup, and then the incremental over top of that. docs/RESTORE talks about how to get data off tapes without the amanda tools. -- Joshua Baker-LePain Department of Biomedical Engineering Duke University
Re: Help With Restore
Hi Kenneth, First off are you sure that you are restoring a tar file. Second try using amrecover. It's an interactive ftp like utility that is very usefull for restoring a single directory or file. chrisj Kenneth Sullivan wrote: I'm trying to restore a directory on one of our servers. First, I want the directory to reside in a temporary location before I move it to its original location. I ran the following command in the temp location: amrestore /dev/nst0 hostname /.1/shares Now I have a file in my temp directory named hostname._.1_shares.20030206.0. I've tried to tar it to expand the contents but the server just sits there, even when running tar overnight. I used: tar xvpS hostname._.1_shares.20030206.0 What am I doing wrong? I need to be able to pull a certain directory out and trash the rest. Any help will be appreciated.
Re: Help With Restore
Kenneth Sullivan wrote: I'm trying to restore a directory on one of our servers. First, I want the directory to reside in a temporary location before I move it to its original location. I ran the following command in the temp location: amrestore /dev/nst0 hostname /.1/shares Now I have a file in my temp directory named hostname._.1_shares.20030206.0. I've tried to tar it to expand the contents but the server just sits there, even when running tar overnight. I used: tar xvpS hostname._.1_shares.20030206.0 What am I doing wrong? I need to be able to pull a certain directory out and trash the rest. Any help will be appreciated. Hi well I tend to use amrecover myself amrecover [ [ -C ] config ] [ -s index-server ] [ -t tape-server ] [ -d tape-device ] (index and tape server tend to be same) then inside amrecover you can set the dir/filesystem you want to recover, where you want it to go and from what date. anyway back to your tar problem. you haven't used the 'f' option to tell tar to use a non default file ie.. tar xvpfS hostname._.1_shares.20030206.0 -- Martin Hepworth Senior Systems Administrator Solid State Logic Ltd +44 (0)1865 842300 ** This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager. This footnote also confirms that this email message has been swept by MIMEsweeper for the presence of computer viruses. www.mimesweeper.com **
Re: Backup and recovery CD
Dear Kevin M. Myer, Once you wrote about Backup and recovery CD: KMM failure, being hacked, etc. I'm wondering if anyone has ever KMM developed something similar for AMANDA. I'm envisioning something KMM like the Red Hat installer, which boots up, lets you drop into a KMM limited shell and lets you partition disks. The AMANDA bootable CD KMM would boot up, let you format disks and give you a limited shell, KMM and then let you run amrestore by connecting to the tape drive (or KMM whatever archival media you are using) of your remote backup KMM server. KMM KMM Just curious if anyone has ever gone down this road and if so, how KMM far did you get? And what's wrong with RedHat rescue mode, since you've mentioned it? :) -- Best regards, Leonid Mamtchenkov, RHCE System Administrator Francoudi Stephanou Ltd.
Re: Newbie question: merge incremental backup to full one
On Wed February 26 2003 11:15, Kablan BOGNINI wrote:
Apologizes for the post because i don't give enough
info.
Here is.
I have make backup on two different tapes say tape A
and B from server_one.
A is holding the full backup and B the incremental
one.
I have another server_two not using amanda where i
want to restore those data ?
What must i do to get all my data(full+incremental)
restore on that server ?
You realise of course this has very little to do with the original
question. (-:
Take the tapes, and prefereably the drive, to the second machine,
even the controller card if need be, and install them to the point
of being operational. Or make sure the network is transparent so
you can access the drive from a shell on the target machine. I'm
not that great on the networking stuff, so I'd move the drive and
controller card. But thats just *my* personal preferences.
Make sure you have the following tools available on the second
machine.
mt
dd
tar (or (shudder) dump)
gzip
The files on the tape are laid out like this
amanda tape ID header (32k)
{
amanda - tar identifying header, may be 32k, I'm not sure
tar'd file represesnting one DLE
}
repeat the {} contents till last DLE is done, at which point an
attempted read of the next header will get you an EOT error.
then
all this assumes the use of the non-rewinding tape device alias.
1. rewind tape
2. dd if=tapedev bs=32k # will show you the tapes name
3. dd if=tapedev bs=32k # will tell you onscreen what the next file
on the tape is, and whether or not you need to pipe it thru gzip,
showing you the actual command line to use.
At this point, go make the target dir for this particular file, and
cd to it.
4. Follow the directions you got from step 3
5. repeat from step 3 until out of data on that tape
6. eject tape and insert 2nd one
7. repeat from 1 to 5.
8. move data to where you want it if the recovery wasn't pointed
there in the first place.
A simple concept, but lots of work. No mistakes allowed unless you
keep track of the file/DLE count so the tape can be rewound, and
then repositioned to the proper starting location with an mt -f
device fsf filenumber where device is say /dev/nst0 (you MUST use
the non-rewinding device for this) and where filenumber might be 26
if thats where you lost track of all the trees in this here forest.
:)
--
Cheers, Gene
AMD [EMAIL PROTECTED] 320M
[EMAIL PROTECTED] 512M
99.23% setiathome rank, not too shabby for a WV hillbilly
UCLA aispdslib amanda
Hi there,
When I compile amanda-2.4.4 on aix 4.3, it gave me following error message:
scsi-aix.c, line 62.10: 1506-296 (S) #include file gscdds.h not found.
scsi-aix.c, line 176.10: 1506-275 (S) Unexpected text scmd encountered.
scsi-aix.c, line 176.3: 1506-045 (S) Undeclared identifier scmd_t.
scsi-aix.c, line 177.3: 1506-275 (S) Unexpected text 'char' encountered.
scsi-aix.c, line 177.8: 1506-045 (S) Undeclared identifier sbyte.
scsi-aix.c, line 178.3: 1506-275 (S) Unexpected text 'int' encountered.
scsi-aix.c, line 178.7: 1506-045 (S) Undeclared identifier Result.
scsi-aix.c, line 179.3: 1506-275 (S) Unexpected text 'int' encountered.
scsi-aix.c, line 179.7: 1506-045 (S) Undeclared identifier isbusy.
scsi-aix.c, line 180.3: 1506-275 (S) Unexpected text 'int' encountered.
scsi-aix.c, line 180.7: 1506-045 (S) Undeclared identifier target.
scsi-aix.c, line 191.7: 1506-045 (S) Undeclared identifier scmd.
scsi-aix.c, line 209.36: 1506-045 (S) Undeclared identifier GSC_CMD.
scsi-aix.c, line 368.22: 1506-280 (W) Function argument assignment between
types const unsigned char* and struct {...}* is not allowed.
make: 1254-004 The error code from the last command is 1
Is there any one know where the gscdds.h is?
Thanks.
Howie
_
Add photos to your messages with MSN 8. Get 2 months FREE*.
http://join.msn.com/?page=features/featuredemail
Re: Moving to production
On Wed February 26 2003 11:51, David Olbersen wrote: Hello all! I read the FAQ about moving from development to production usage. http://amanda.sourceforge.net/cgi-bin/fom?_highlightWords=tapelis tfile=34 That answer doesn't mention what to do with the 'index' directory -- do I need to remove this or leave it as it is? Does it make a difference now that I have new tapes? Amanda has some housekeeping functions, and those indices that go out of date because the tape has been re-used will normally be deleted automaticly too. Shes a good girl, carries her own vacuum cleaner to handle the leftovers. :-) -- Cheers, Gene AMD [EMAIL PROTECTED] 320M [EMAIL PROTECTED] 512M 99.23% setiathome rank, not too shabby for a WV hillbilly
Two more amanda questions
1) This morning, my amanda report said:
These dumps were to tape Indyme010.
The next tape Amanda expects to use is: a new tape.
Instead of The next tape Amanda expects to use is: Indyme011.
Here's the tapelist:
20030226 Indyme010 reuse
20030225 Indyme009 reuse
20030221 Indyme008 reuse
20030220 Indyme007 reuse
20030219 Indyme006 reuse
20030218 Indyme005 reuse
20030215 Indyme004 reuse
20030214 Indyme003 reuse
20030213 Indyme002 reuse
20030212 Indyme001 reuse
0 Indyme013 reuse
0 Indyme012 reuse
0 Indyme011 reuse
There was a 014 and 015 in there, but I deleted them so, afetr this
week, it would go back to using 001 on Monday.
2) Also, my report gave me another STRANGE backup, caused in part by
stuff in /tmp and the holdingdisk changing. But:
[EMAIL PROTECTED] root]# cat /usr/local/lib/amanda/exclude.gtar
/tmp
/hold
define dumptype root-tar {
global
program GNUTAR
comment root partitions dumped with tar
compress none
index
exclude list /usr/local/lib/amanda/exclude.gtar
priority low
}
define dumptype comp-root-tar {
root-tar
comment Root partitions with compression
compress client fast
}
[EMAIL PROTECTED] root]# grep backup /etc/amanda/DailySet1/disklist
backup /dev/hda1 comp-root-tar
What am I doing wrong to get /tmp and /hold excluded?
--
John Oliver, CCNAhttp://www.john-oliver.net/
Linux/UNIX/network consulting http://www.john-oliver.net/resume/
*** sendmail, Apache, ftp, DNS, spam filtering ***
Colocation, T1s, web/email/ftp hosting
Re: Amanda Backup Strategy
On Wed February 26 2003 11:16, Ralph Bearpark wrote: Yep. Amanda gets estimates for all the DLEs at multiple incremental levels, and figures out the best thing to do within the tapelength it has. Even to the extent of not doing any sort of backup at all on some or many of the DLEs? If so, how on earth does it decide which? Amanda shouldn't have to deal with that, so the normal startup procedure is to only give amanda enough acive DLE's to fill a tape each night until all have been done, thereby giving amanda some help on the initial balance settings. But, if the tapetype is properly set, I don't *think* amanda will take a chance on hitting EOT as it will delay enough to assure it won't hit EOT. I've seen that in my email from amanda a time or 2. (If this is all documented somewhere, please tell me where and I'll stop asking my stupid questions! :-)) The only really stupid question is the one not asked. ;-) -- Cheers, Gene AMD [EMAIL PROTECTED] 320M [EMAIL PROTECTED] 512M 99.23% setiathome rank, not too shabby for a WV hillbilly
RE: Backup and recovery CD
Leonid, I have been trying to do a complete restore on a Red Hat box for some time now. I haven't been able to figure out a way to access my SCSI tape drive from Red Hat rescue mode. Perhaps I'm missing something? -Original Message- From: Leonid Mamchenkov [mailto:[EMAIL PROTECTED] Sent: Wednesday, February 26, 2003 11:50 AM To: [EMAIL PROTECTED] Subject: Re: Backup and recovery CD Dear Kevin M. Myer, Once you wrote about Backup and recovery CD: KMM failure, being hacked, etc. I'm wondering if anyone has ever KMM developed something similar for AMANDA. I'm envisioning something KMM like the Red Hat installer, which boots up, lets you drop into a KMM limited shell and lets you partition disks. The AMANDA bootable CD KMM would boot up, let you format disks and give you a limited shell, KMM and then let you run amrestore by connecting to the tape drive (or KMM whatever archival media you are using) of your remote backup KMM server. KMM KMM Just curious if anyone has ever gone down this road and if so, how KMM far did you get? And what's wrong with RedHat rescue mode, since you've mentioned it? :) -- Best regards, Leonid Mamtchenkov, RHCE System Administrator Francoudi Stephanou Ltd.
RE: Help With Restore
I'm trying to restore a directory on one of our servers. First, I want the directory to reside in a temporary location before I move it to its original location. I ran the following command in the temp location: amrestore /dev/nst0 hostname /.1/shares Now I have a file in my temp directory named hostname._.1_shares.20030206.0. I've tried to tar it to expand the contents but the server just sits there, even when running tar overnight. I used: tar xvpS hostname._.1_shares.20030206.0 What am I doing wrong? I need to be able to pull a certain directory out and trash the rest. Any help will be appreciated. It's been my understanding that you can only use amrecover for individual files, and amrestore for entire filesystems. (I could be wrong) And in order to use amrecover, you have to have your dumptype index parameter set to yes somewhere in your config. Also, you need to have amindexd and amidxtaped installed. try: # amrecover configname you should see come connection information and then an amrecover prompt. from here you can set the date, the host, the disk. You can cd and ls within the amrecover session and then add the file, once it's added then extract. NOTE: wherever you are at the time in which you amrecover is where the file will restore. Hope this helps...good luck.
Re: Help With Restore
On Wed, Feb 26, 2003 at 02:10:31PM -0600, Rebecca Pakish Crum wrote: It's been my understanding that you can only use amrecover for individual files, amrecover is quite happy to recover the top directory (and thereby all it's content). and amrestore for entire filesystems. (I could be wrong) If you're using dump/ufsdump/xfsdump, amrestore is a bit more efficient for a full restore, because you can 'ufsrestore r' instead of 'ufsrestore i'. And in order to use amrecover, you have to have your dumptype index parameter set to yes somewhere in your config. Also, you need to have amindexd and amidxtaped installed. Yes. -- Jay Lessert [EMAIL PROTECTED] Accelerant Networks Inc. (voice)1.503.439.3461 Beaverton OR, USA(fax)1.503.466.9472
AMRecover
I still cant figure out why I dont see files from incremental backup. I see only files from full backup with amrecover ls command. Im using file based backup with some slots, so each backup data goes to another folder. I can see files from the incremental backup in data and also in index files but with amrecover I cant list them. Any suggestions? Regards, Gregor
Re: Two more amanda questions
On Wed, 26 Feb 2003 at 11:33am, John Oliver wrote These dumps were to tape Indyme010. The next tape Amanda expects to use is: a new tape. Instead of The next tape Amanda expects to use is: Indyme011. Here's the tapelist: 20030226 Indyme010 reuse 20030225 Indyme009 reuse 20030221 Indyme008 reuse 20030220 Indyme007 reuse 20030219 Indyme006 reuse 20030218 Indyme005 reuse 20030215 Indyme004 reuse 20030214 Indyme003 reuse 20030213 Indyme002 reuse 20030212 Indyme001 reuse 0 Indyme013 reuse 0 Indyme012 reuse 0 Indyme011 reuse That's normal -- amanda can't read your mind, and doesn't know that you intend on using Indyme011 next. Once you've used tapecycle worth of tapes once, amanda will start asking for them in the same order in which they were first used. 2) Also, my report gave me another STRANGE backup, caused in part by stuff in /tmp and the holdingdisk changing. But: [EMAIL PROTECTED] root]# grep backup /etc/amanda/DailySet1/disklist backup /dev/hda1 comp-root-tar What am I doing wrong to get /tmp and /hold excluded? I think you want ./tmp and ./hold in there -- relative paths. -- Joshua Baker-LePain Department of Biomedical Engineering Duke University
amrecover: Connection refused
Hi, I've configured tapeless backup, and trying to verify the data by running amrecover. It says Connection refused I had port in /etc/services, and xinetd enabled; and also an entry lnx200.gibbons.com root in .amandahosts Server is itself client, and recovery is done on it. Here's the complete message (Amanda 2.4.3) AMRECOVER Version 2.4.3. Contacting server on lnx200.gibbons.com ... amrecover: cannot connect to lnx200.gibbons.com: Connection refused Can anyone help me resolve this? Thanks,
Re: Two more amanda questions
On Wed, Feb 26, 2003 at 03:54:16PM -0500, Joshua Baker-LePain wrote: 20030226 Indyme010 reuse 20030225 Indyme009 reuse 20030221 Indyme008 reuse 20030220 Indyme007 reuse 20030219 Indyme006 reuse 20030218 Indyme005 reuse 20030215 Indyme004 reuse 20030214 Indyme003 reuse 20030213 Indyme002 reuse 20030212 Indyme001 reuse 0 Indyme013 reuse 0 Indyme012 reuse 0 Indyme011 reuse That's normal -- amanda can't read your mind, and doesn't know that you intend on using Indyme011 next. Once you've used tapecycle worth of tapes once, amanda will start asking for them in the same order in which they were first used. I thought amanda would just read from the bottom of the list. When it was done with tape 009, 010 was at the bottom of the list and was used. Now 010 is done, and 011 is at the bottom of the list. Why isn't it just taking that, instead of trying to read my mind? :-) dumpcycle 1 week runspercycle 5 tapecycle 15 tapes What am I doing wrong to get /tmp and /hold excluded? I think you want ./tmp and ./hold in there -- relative paths. Ahhh! Thanks... -- John Oliver, CCNAhttp://www.john-oliver.net/ Linux/UNIX/network consulting http://www.john-oliver.net/resume/ *** sendmail, Apache, ftp, DNS, spam filtering *** Colocation, T1s, web/email/ftp hosting
Re: amrecover: Connection refused
I found it. The installation set the user of amandaidx and amidxtape to amanda, but I dont know why Amanda requires root to run it, which conflicts with the default setting. bao wrote: Hi, I've configured tapeless backup, and trying to verify the data by running amrecover. It says Connection refused I had port in /etc/services, and xinetd enabled; and also an entry lnx200.gibbons.comroot in .amandahosts Server is itself client, and recovery is done on it. Here's the complete message (Amanda 2.4.3) AMRECOVER Version 2.4.3. Contacting server on lnx200.gibbons.com ... amrecover: cannot connect to lnx200.gibbons.com: Connection refused Can anyone help me resolve this? Thanks,
RE: Two more amanda questions
1) This morning, my amanda report said: These dumps were to tape Indyme010. The next tape Amanda expects to use is: a new tape. Instead of The next tape Amanda expects to use is: Indyme011. Here's the tapelist: 20030226 Indyme010 reuse 20030225 Indyme009 reuse 20030221 Indyme008 reuse 20030220 Indyme007 reuse 20030219 Indyme006 reuse 20030218 Indyme005 reuse 20030215 Indyme004 reuse 20030214 Indyme003 reuse 20030213 Indyme002 reuse 20030212 Indyme001 reuse 0 Indyme013 reuse 0 Indyme012 reuse 0 Indyme011 reuse There was a 014 and 015 in there, but I deleted them so, afetr this week, it would go back to using 001 on Monday. snip When did you label 011-013? Not to panic...those tapes haven't been used yet...as long as your tapecycle lists at least 13 tapes, it will take Indyme011 no problem. Then once it's been used and listed in the tapelist with a date, amanda will expect it after Indyme010. Pop it in and run amcheck...you'll see it's okay.
RE: Amanda Backup Strategy
Someone recently recommended Unix Backup Recovery by W. Curtis Preston to me. I am just now wading through the amanda section so I cannot tell you if it's any good. But, there certainly is a lot of material there! -Original Message- From: Gene Heskett [mailto:[EMAIL PROTECTED] Sent: Wednesday, February 26, 2003 12:29 PM To: Ralph Bearpark; [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Subject: Re: Amanda Backup Strategy On Wed February 26 2003 11:16, Ralph Bearpark wrote: Yep. Amanda gets estimates for all the DLEs at multiple incremental levels, and figures out the best thing to do within the tapelength it has. Even to the extent of not doing any sort of backup at all on some or many of the DLEs? If so, how on earth does it decide which? Amanda shouldn't have to deal with that, so the normal startup procedure is to only give amanda enough acive DLE's to fill a tape each night until all have been done, thereby giving amanda some help on the initial balance settings. But, if the tapetype is properly set, I don't *think* amanda will take a chance on hitting EOT as it will delay enough to assure it won't hit EOT. I've seen that in my email from amanda a time or 2. (If this is all documented somewhere, please tell me where and I'll stop asking my stupid questions! :-)) The only really stupid question is the one not asked. ;-) -- Cheers, Gene AMD [EMAIL PROTECTED] 320M [EMAIL PROTECTED] 512M 99.23% setiathome rank, not too shabby for a WV hillbilly
Re: amrecover: Connection refused
On Wed, 26 Feb 2003, bao wrote: I found it. The installation set the user of amandaidx and amidxtape to amanda, but I dont know why Amanda requires root to run it, which conflicts with the default setting. Amanda run amandaidx and amidxtape under the user amanda. but it s the server part amrecover is the client part and then should run under the identity of root. bao wrote: Hi, I've configured tapeless backup, and trying to verify the data by running amrecover. It says Connection refused I had port in /etc/services, and xinetd enabled; and also an entry lnx200.gibbons.comroot in .amandahosts Server is itself client, and recovery is done on it. Here's the complete message (Amanda 2.4.3) AMRECOVER Version 2.4.3. Contacting server on lnx200.gibbons.com ... amrecover: cannot connect to lnx200.gibbons.com: Connection refused Can anyone help me resolve this? Thanks, -- Eric Doutreleau I.N.T | Tel : +33 (0) 160764687 9 rue Charles Fourier | Fax : +33 (0) 160764321 91011 Evry France | email : [EMAIL PROTECTED]
Re: Two more amanda questions
On Wed February 26 2003 14:33, John Oliver wrote:
1) This morning, my amanda report said:
These dumps were to tape Indyme010.
The next tape Amanda expects to use is: a new tape.
Instead of The next tape Amanda expects to use is: Indyme011.
Here's the tapelist:
20030226 Indyme010 reuse
20030225 Indyme009 reuse
20030221 Indyme008 reuse
20030220 Indyme007 reuse
20030219 Indyme006 reuse
20030218 Indyme005 reuse
20030215 Indyme004 reuse
20030214 Indyme003 reuse
20030213 Indyme002 reuse
20030212 Indyme001 reuse
0 Indyme013 reuse
0 Indyme012 reuse
0 Indyme011 reuse
There was a 014 and 015 in there, but I deleted them so, afetr
this week, it would go back to using 001 on Monday.
2) Also, my report gave me another STRANGE backup, caused in part
by stuff in /tmp and the holdingdisk changing. But:
[EMAIL PROTECTED] root]# cat /usr/local/lib/amanda/exclude.gtar
/tmp
/hold
exclude lists were, the last time I used them, to be stated in
relative format. If you had entered ./tmp and ./hold (note the
leading dot to anchor it to the present directory) then *any* tmp
dir and *any* hold dir will have been skipped, including the
subdirs of any such named dir.
But please read the docs, this may have been changed since the last
time I used this feature. As the stuff I figure is loseable in
/tmp, then I just don't put it in my disklist, and of course I do
not have a / entry in my disklist as that would take about 6 of
my 4 gig tapes to cover in one swell foop. I can regenerate it if
I need it bad enough.
define dumptype root-tar {
global
program GNUTAR
comment root partitions dumped with tar
compress none
index
exclude list /usr/local/lib/amanda/exclude.gtar
priority low
}
define dumptype comp-root-tar {
root-tar
comment Root partitions with compression
compress client fast
}
[EMAIL PROTECTED] root]# grep backup /etc/amanda/DailySet1/disklist
backup /dev/hda1 comp-root-tar
What am I doing wrong to get /tmp and /hold excluded?
--
Cheers, Gene
AMD [EMAIL PROTECTED] 320M
[EMAIL PROTECTED] 512M
99.23% setiathome rank, not too shabby for a WV hillbilly
Re: Backup and recovery CD
On Wed February 26 2003 15:06, Seth, Wayne (Contractor) wrote: Leonid, I have been trying to do a complete restore on a Red Hat box for some time now. I haven't been able to figure out a way to access my SCSI tape drive from Red Hat rescue mode. Perhaps I'm missing something? -Original Message- From: Leonid Mamchenkov [mailto:[EMAIL PROTECTED] Sent: Wednesday, February 26, 2003 11:50 AM To: [EMAIL PROTECTED] Subject: Re: Backup and recovery CD Dear Kevin M. Myer, Once you wrote about Backup and recovery CD: KMM failure, being hacked, etc. I'm wondering if anyone has ever KMM developed something similar for AMANDA. I'm envisioning something KMM like the Red Hat installer, which boots up, lets you drop into a KMM limited shell and lets you partition disks. The AMANDA bootable CD KMM would boot up, let you format disks and give you a limited shell, KMM and then let you run amrestore by connecting to the tape drive (or KMM whatever archival media you are using) of your remote backup KMM server. KMM KMM Just curious if anyone has ever gone down this road and if so, how KMM far did you get? And what's wrong with RedHat rescue mode, since you've mentioned it? :) I can report one problem with that, if he is booting from the updated kernel that up2date might have installed. The updated kernel did *not* come accompanied by an updated initrd, and the updated kernel is useless for such booting because theres no matching version numbered initrd he can load his scsi drivers from. I know, I let it install it on me, and it over-wrote my grub.conf rendering the rest of my custom kernels inactive till I fixed that by giving grub a custom command line that did point to a valid for my system kernel, and used that to restore my grub.conf. I have not allowed up2date to touch my kernel installs since that fiasco. A rather heated message to RH at the time was not replied to. There are many things up2date is good for, but putzing with your kernel install isn't one of them. -- Cheers, Gene AMD [EMAIL PROTECTED] 320M [EMAIL PROTECTED] 512M 99.23% setiathome rank, not too shabby for a WV hillbilly
Re: Amanda Backup Strategy
On Wed February 26 2003 17:22, Seth, Wayne (Contractor) wrote: Someone recently recommended Unix Backup Recovery by W. Curtis Preston to me. I am just now wading through the amanda section so I cannot tell you if it's any good. But, there certainly is a lot of material there! Yes, and your quoting is broken, making it look as if I wrote both you message above, and my message below. Please adjust your email agent. -Original Message- From: Gene Heskett [mailto:[EMAIL PROTECTED] Sent: Wednesday, February 26, 2003 12:29 PM To: Ralph Bearpark; [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Subject: Re: Amanda Backup Strategy On Wed February 26 2003 11:16, Ralph Bearpark wrote: Yep. Amanda gets estimates for all the DLEs at multiple incremental levels, and figures out the best thing to do within the tapelength it has. Even to the extent of not doing any sort of backup at all on some or many of the DLEs? If so, how on earth does it decide which? Amanda shouldn't have to deal with that, so the normal startup procedure is to only give amanda enough acive DLE's to fill a tape each night until all have been done, thereby giving amanda some help on the initial balance settings. But, if the tapetype is properly set, I don't *think* amanda will take a chance on hitting EOT as it will delay enough to assure it won't hit EOT. I've seen that in my email from amanda a time or 2. (If this is all documented somewhere, please tell me where and I'll stop asking my stupid questions! :-)) The only really stupid question is the one not asked. ;-) -- Cheers, Gene AMD [EMAIL PROTECTED] 320M [EMAIL PROTECTED] 512M 99.23% setiathome rank, not too shabby for a WV hillbilly
Re: Two more amanda questions
On Wed, Feb 26, 2003 at 01:38:46PM -0800, John Oliver wrote: I thought amanda would just read from the bottom of the list. Nope, amanda goes by date first. List position is only significant when 2 or more tapes have the same last-used date. When it was done with tape 009, 010 was at the bottom of the list and was used. Now 010 is done, and 011 is at the bottom of the list. Why isn't it just taking that, instead of trying to read my mind? :-) Just a guess, but I'd say it's because 0 is not a valid date, so those lines are being ignored completely.
Re: Two more amanda questions
On Wed, Feb 26, 2003 at 05:26:59PM -0600, Dave Sherohman wrote: On Wed, Feb 26, 2003 at 01:38:46PM -0800, John Oliver wrote: I thought amanda would just read from the bottom of the list. Nope, amanda goes by date first. List position is only significant when 2 or more tapes have the same last-used date. When it was done with tape 009, 010 was at the bottom of the list and was used. Now 010 is done, and 011 is at the bottom of the list. Why isn't it just taking that, instead of trying to read my mind? :-) Just a guess, but I'd say it's because 0 is not a valid date, so those lines are being ignored completely. Those lines are what results when you amlabel a tape. If your surmisation was correct, amanda would never use any tapes. -- John Oliver, CCNAhttp://www.john-oliver.net/ Linux/UNIX/network consulting http://www.john-oliver.net/resume/ *** sendmail, Apache, ftp, DNS, spam filtering *** Colocation, T1s, web/email/ftp hosting
Re: Backup and recovery CD
Dear Gene Heskett, Once you wrote about Re: Backup and recovery CD: GH And what's wrong with RedHat rescue mode, since you've mentioned GH it? :) GH GH I can report one problem with that, if he is booting from the GH updated kernel that up2date might have installed. The updated GH kernel did *not* come accompanied by an updated initrd, and the GH updated kernel is useless for such booting because theres no GH matching version numbered initrd he can load his scsi drivers from. initrd, IIRC, never comes with the kernel. Instead, I beleive, it is generated by mkinitrd during the installation of the RPM. Anyway, the rescue mode is not even affected by these problem, since one is using a floppy or CD-ROM to go to rescue mode, not a broken installation. :) GH I know, I let it install it on me, and it over-wrote my grub.conf GH rendering the rest of my custom kernels inactive till I fixed that GH by giving grub a custom command line that did point to a valid for GH my system kernel, and used that to restore my grub.conf. I have GH not allowed up2date to touch my kernel installs since that fiasco. GH GH A rather heated message to RH at the time was not replied to. GH GH There are many things up2date is good for, but putzing with your GH kernel install isn't one of them. You might want to try a newer version. It work much better then it used to with in times of RedHat 6.x, if you know what I mean. :) -- Best regards, Leonid Mamtchenkov, RHCE System Administrator Francoudi Stephanou Ltd.
Re: Backup and recovery CD
Dear Seth, Wayne (Contractor), Once you wrote about RE: Backup and recovery CD: SWC I have been trying to do a complete restore on a Red Hat box for some time SWC now. I haven't been able to figure out a way to access my SCSI tape drive SWC from Red Hat rescue mode. Perhaps I'm missing something? Hmm, interesting. I, myself, never had to use SCSI tape on the recoring machine. Most of the times, it's a network restore. But, IIRC, SCSI tape drives are perfectly detectable during the boot time and tar is a part of the rescue mode. SWC Dear Kevin M. Myer, SWC SWC Once you wrote about Backup and recovery CD: SWC KMM failure, being hacked, etc. I'm wondering if anyone has ever SWC KMM developed something similar for AMANDA. I'm envisioning something SWC KMM like the Red Hat installer, which boots up, lets you drop into a SWC KMM limited shell and lets you partition disks. The AMANDA bootable CD SWC KMM would boot up, let you format disks and give you a limited shell, SWC KMM and then let you run amrestore by connecting to the tape drive (or SWC KMM whatever archival media you are using) of your remote backup SWC KMM server. SWC KMM SWC KMM Just curious if anyone has ever gone down this road and if so, how SWC KMM far did you get? SWC SWC And what's wrong with RedHat rescue mode, since you've mentioned it? :) -- Best regards, Leonid Mamtchenkov, RHCE System Administrator Francoudi Stephanou Ltd.
