Re: Dump command on Debian
well, when using a non-standard filesystem like jfs on linux, i don't think you have much choice but to use gtar, which doesn't depend on the filesystem used. but anyone, feel free to correct me if i'm wrong about this.. -rodi. On Mon, 2004-07-12 at 22:28, John Bossert wrote: > Gtar works fine (and I use it for my Dailies.) > > Here, I'm trying to establish a "bare metal" restore process. With > Solaris, if _really bad things_ happen, I can take my ufs dumps and > rebuild a machine fairly directly. > > Is the "best practice" in the Debian world to just use gtar for this? > > How do others manage this? Thx. > > -john > > R.M. Evers wrote: > > > john, > > > > you say gtar and dump work on the sun boxes, but only state that dump > > fails on the debian boxes (which is probably correct, since it only > > dumps ext2 filesystems). did you try gtar? it should work.. :-) > > > > regards, > > -rodi. > > > > > > On Sat, 2004-07-10 at 17:07, John Bossert wrote: > > > >>Adding a pair of Debian(woody) clients to a Solaris shop... Amanda is > >>working just fine with both gtar and dump on the Sun boxes. > >> > >>Trying to dump filesystems, failed. Discovered that there's no "dump" > >>on the debian boxes... > >> > >>All the filesystems are jfs, yet the "dump" package I find at debian.org > >>(0.4b27-4) appears oriented towards ext2. > >> > >>What package do I want to install on these clients to play nice with > >>Amanda? Thx.
Re: Dump command on Debian
john, you say gtar and dump work on the sun boxes, but only state that dump fails on the debian boxes (which is probably correct, since it only dumps ext2 filesystems). did you try gtar? it should work.. :-) regards, -rodi. On Sat, 2004-07-10 at 17:07, John Bossert wrote: > Adding a pair of Debian(woody) clients to a Solaris shop... Amanda is > working just fine with both gtar and dump on the Sun boxes. > > Trying to dump filesystems, failed. Discovered that there's no "dump" > on the debian boxes... > > All the filesystems are jfs, yet the "dump" package I find at debian.org > (0.4b27-4) appears oriented towards ext2. > > What package do I want to install on these clients to play nice with > Amanda? Thx.
Re: A server rebuild question
probably you'll have to run "amcheck -a YourConfig"? regards, -rodi. On Wed, 2004-07-07 at 11:18, Gavin Henry wrote: > Hi guys, > > I am falling in love more and more with amanda now. > > Got a new server and all set. > > Just a quick one. I remember that before, when running amcheck, I used > to get an e-mail after. I don't seem to be getting this, but I do get > my completed backup email from amdump like normal. > > Gavin. > > PS. Are there any better amanda documentation resources out there? I > feel the restore one needs to be redone/added too. > > May I? > > P.P.S> it will be posted also on may page at > http://FedoraNEWS.ORG/ghenry when I'm done, as I have been meaning to > do the restore howto for a while now. >
Re: Strange DNS lookup problems ... I think ...
hi geoff, i've had a similar problem about a week ago. look at this report-snippet (i've taken out the hosts and ip's): -- These dumps were to tape daily09. The next tape Amanda expects to use is: a new tape. The next new tape already labelled is: daily10. FAILURE AND STRANGE DUMP SUMMARY: nptn.some.net /var/lib/mysql_backup lev 0 FAILED [nptn.some.net: [addr 1.2.3.4: hostname lookup failed]] nptn.some.net /etc lev 0 FAILED [nptn.some.net: [addr 1.2.3.4: hostname lookup failed]] nptn.some.net /home lev 0 FAILED [nptn.some.net: [addr 1.2.3.4: hostname lookup failed]] -- also in my case, it's very unlikely that it's a dns issue. we run our own nameservers, and none have had any downtime. forward and reverse is also properly configured.. i have no answer to your question, but in my case the problem seems to have fixxed itself.. the next day this host was backed up properly, and i've had no problem since. maybe amanda's dns lookup process is a bit flakey? regards, -Rodi Evers On Fri, 2004-03-05 at 04:44, Geoff Swavley wrote: > hi All, > > I run 25 schedules per night, and only 1 ... this one ... gives me some > heart-ache. I know it must be something other than DNS but all of the > error messages are trying to send me down that garden path. The amanda > host machine is "windamere", and schedule6 backs up "intrap". I can log onto > both machines and easily do forward and reverse lookups on both hosts, so I know > > there is no DNS issue. > --- > windamere[26]% nslookup intrap > Server: gps.dlwc.nsw.gov.au > Address: 172.24.16.91 > > Name:intrap.dlwc.nsw.gov.au > Address: 172.24.16.32 > > windamere[27]% nslookup 172.24.16.32 > Server: gps.dlwc.nsw.gov.au > Address: 172.24.16.91 > > Name:intrap.dlwc.nsw.gov.au > Address: 172.24.16.32 > > windamere[28]% > -- > intrap[1]# nslookup windamere > Server: gps.dlwc.nsw.gov.au > Address: 172.24.16.91 > > Name:windamere.dlwc.nsw.gov.au > Address: 172.24.16.86 > > intrap[2]# nslookup 172.24.16.86 > Server: gps.dlwc.nsw.gov.au > Address: 172.24.16.91 > > Name:windamere.dlwc.nsw.gov.au > Address: 172.24.16.86 > > intrap[3]# > > > Here's the output from the schedule 6 dump - planner is NOT HAPPY > about something any ideas what's going wrong???: > > Subject:schedule6 AMANDA MAIL REPORT FOR March 5, 2004 >Date: Fri, 5 Mar 2004 00:15:07 +1100 (EST) >From:Amanda Archiving Server <[EMAIL PROTECTED]> > To: [EMAIL PROTECTED] > > These dumps were to tape schedule6-THU. > The next tape Amanda expects to use is: schedule6-MON. > > FAILURE AND STRANGE DUMP SUMMARY: > planner: ERROR intrap: [addr 172.24.16.86: hostname lookup failed] > intrap / RESULTS MISSING > intrap /appl RESULTS MISSING > intrap /export/home RESULTS MISSING > intrap /htdocs RESULTS MISSING > intrap /opt RESULTS MISSING > intrap /opt/apache RESULTS MISSING > intrap /u01 RESULTS MISSING > intrap /u02 RESULTS MISSING > > > STATISTICS: > Total Full Daily > > Estimate Time (hrs:min)0:00 > Run Time (hrs:min) 0:00 > Dump Time (hrs:min)0:00 0:00 0:00 > Output Size (meg) 0.00.00.0 > Original Size (meg) 0.00.00.0 > Avg Compressed Size (%) -- -- -- > Filesystems Dumped0 0 0 > Avg Dump Rate (k/s) -- -- -- > > Tape Time (hrs:min)0:00 0:00 0:00 > Tape Size (meg) 0.00.00.0 > Tape Used (%) 0.00.00.0 > Filesystems Taped 0 0 0 > Avg Tp Write Rate (k/s) -- -- -- > > USAGE BY TAPE: > Label Time Size %Nb > schedule6-THU 0:00 0.00.0 0 > > > NOTES: > planner: tapecycle (4) <= runspercycle (4) > driver: WARNING: got empty schedule from planner > taper: tape schedule6-THU kb 0 fm 0 [OK] > > > DUMP SUMMARY: > DUMPER STATSTAPER STATS > HOSTNAME DISKL ORIG-KB OUT-KB COMP% MMM:SS KB/s MMM:SS KB/s > -- - > intrap / MISSING -- > intrap /appl MISSING -- > intrap -xport/home MISSING -- > intrap /htdocs MISSING -- > intrap /opt MISSING ---
Re: amrecovering extremely slow
On Fri, 2004-02-06 at 17:37, Paul Bijnens wrote: > (If you have strong programming skills, you could add an utility > to gnutar, that extracts only the wanted files, packs them again > on the fly to send them over the network, where another gnutar program > accepts only the useful bytes. And while you have built your expertise, > you could do the same for all the dump variants. :-) ) I'll start right away :-p Thanks Paul, -Rodi.
Re: amrecovering extremely slow
On Fri, 2004-02-06 at 17:14, Jon LaBadie wrote: > I'm assuming you were recovering something from /home. > > Yes I think the entire dump file would have to be passed to the > client for extraction. It is nice that you are using similar OS's > and backup programs (probably gnutar) on both tapeserver and client. > > But suppose your tapeserver were say linux on an Intel processor > and the client were an HP-UX system, using vxdump/vxrestore, on > an HP PA-RISC processor. There is no way the tapeserver could run > the restore program, it has to send the dump image to the client > that has the extraction program. Jon & Stefan, thank you so much for your fast replies and clearing this up for me. Indeed I always use Linux and tar, forgetting some people actually use other stuff :-) Shame on me. From now on I will just recover my backups on the backupserver itself, and then transfer it to my workstation. Lot faster (and probably safer) that way. Thanks again. -Rodi.
amrecovering extremely slow
amanda-users, after the last problem in my setup (the chg-disk not working, now working fine :-)), i've run into another one. although i don't know whether it's really a problem. i've done the following: 1.- used my brand spanking new amanda backupserver (amanda-2.4.4p2, self-compiled) to create a backup of server1 (amanda-2.4.2p2, debian/stable-packaged) using the file: driver. this went fine. 2.- tried to amrecover from the backupserver itself. although the extract-phase sometimes paused for a bit, this also went fine. 3.- tried to amrecover from workstation1 (amanda-2.4.4p2, self-compiled), and this went very slwww... what happened is this. i started the amrecover (-s backupserver -t backupserver) session to the backupserver. set the host, disk, added a homedir, set the tape to chg-disk (my 'amrecover_changer') and then hit extract. then the directory-structure came in. but after that, i had to wait a long time for the actual files to come in (really really long). and during this time, i had a LOT of network traffic between my backupserver and my workstation (i mean like 850kB/s continuously for 5-10 minutes). it's almost as if the whole /home was sent to my workstation or something. is this normal behaviour, or does anybody recognize it? or have i possibly compiled or ./configure'd amanda completely wrong? thanks & regards, -rodi.
Re: amrecover with file:, again.....
Compiling amanda-2.4.4p2 myself did the trick. I adapted Marc's config for my site, et voila! Thanks again. Regards, -Rodi On Wed, 2004-02-04 at 09:56, R.M. Evers wrote: > Thanks for your reply Marc, but alas, still no-go :-( The > amidxtaped.xxx.debug file shows the following info: > > ... > amrestore: could not stat chg-disk: No such file or directory > amidxtaped: time 0.063: amrestore terminated normally with status: 2 > amidxtaped: could not stat chg-disk: No such file or directory > amidxtaped: time 0.064: could not stat chg-disk: No such file or > directory > ... > > So it seems that the chg-disk script can't be found, which sits in > /usr/lib/amanda along with the other scripts.. Maybe something went > wrong with my way of installing and patching amanda. I'll fetch the > latest stable version from amanda.org, and try compiling it myself. Back > to the drawing board it is > > Regards, > -Rodi. > > > On Tue, 2004-02-03 at 16:41, Marc Langlois wrote: > > Hi Rodi, > > > > The "amrecover: Can't read file header" message happened to me when the > > chg-disk script wasn't being called, and amrecover was trying to read > > from an empty tape slot directory i.e. the "data" sym-link was wrong. > > > > My amanda.conf has these entries: > > > > tpchanger "chg-disk" > > tapedev "file:/s4/amanda_tapes" > > rawtapedev "file:/s4/amanda_tapes" > > changerfile "/usr/local/etc/amanda/clst01/changer" > > changerdev "/dev/null" > > amrecover_changer "chg-disk" > > > > In amrecover, I do: > > > > amrecover> settape chg-disk > > amrecover> extract > > ... > > > > and it works fine. Doing an "ls -l" on the tape slot directory > > (/s4/amanda_tapes for me) shows that the "data" sym-link is changed to > > point to the correct tape slot before each extract begins. > > > > The value of "amrecover_changer" seems to be arbitrary - as long as the > > same string is used in amrecover, then it will use the value of > > "tpchanger" (AFAIK). > > > > Hope this helps, > > Marc Langlois. > > > > On Tue, 2004-02-03 at 06:34, R.M. Evers wrote: > > > hi amanda-users, > > > > > > sorry to bring this up again, but i'm also having problems with using > > > amrecover with the file: driver.. i have read the recent thread, but i > > > cannot get it to work properly. > > snip... > > > > > > then, i tried to recover some directory using amrecover. at the final > > > 'extract'-point, i receive the error: > > > > > > -- > > > Extracting files using tape drive daily01 on host bu2. > > > Load tape daily02 now > > > Continue [?/Y/n/s/t]? > > > EOF, check amidxtaped..debug file on bu2. > > > amrecover: short block 0 bytes > > > UNKNOWN file > > > amrecover: Can't read file header > > > extract_list - child returned non-zero status: 1 > > > -- > > > > > > the recover works fine when i issue an "amtape DailySet1 label daily02" > > > on the backup server, but hey, i'm lazy.. > > > > > > i've tried different values for the amrecover_changer parameter > > > mentioned in the thread last month, but to no avail.. right now, here's > > > my config: > > > > > > -- > > > tpchanger "chg-disk" > > > changerfile "/etc/amanda/DailySet1/changer" > > > tapedev "file:/var/backups/amanda-tapes/DailySet1" > > > amrecover_changer "file:/var/backups/amanda-tapes/DailySet1" > > > -- > > > > > > i read the docs, googled like there's no tomorrow, but i can't seem to > > > get this right. what am i missing here? > > > > > > thanks & regards, > > > -rodi.
Re: amrecover with file:, again.....
Thanks for your reply Marc, but alas, still no-go :-( The amidxtaped.xxx.debug file shows the following info: ... amrestore: could not stat chg-disk: No such file or directory amidxtaped: time 0.063: amrestore terminated normally with status: 2 amidxtaped: could not stat chg-disk: No such file or directory amidxtaped: time 0.064: could not stat chg-disk: No such file or directory ... So it seems that the chg-disk script can't be found, which sits in /usr/lib/amanda along with the other scripts.. Maybe something went wrong with my way of installing and patching amanda. I'll fetch the latest stable version from amanda.org, and try compiling it myself. Back to the drawing board it is Regards, -Rodi. On Tue, 2004-02-03 at 16:41, Marc Langlois wrote: > Hi Rodi, > > The "amrecover: Can't read file header" message happened to me when the > chg-disk script wasn't being called, and amrecover was trying to read > from an empty tape slot directory i.e. the "data" sym-link was wrong. > > My amanda.conf has these entries: > > tpchanger "chg-disk" > tapedev "file:/s4/amanda_tapes" > rawtapedev "file:/s4/amanda_tapes" > changerfile "/usr/local/etc/amanda/clst01/changer" > changerdev "/dev/null" > amrecover_changer "chg-disk" > > In amrecover, I do: > > amrecover> settape chg-disk > amrecover> extract > ... > > and it works fine. Doing an "ls -l" on the tape slot directory > (/s4/amanda_tapes for me) shows that the "data" sym-link is changed to > point to the correct tape slot before each extract begins. > > The value of "amrecover_changer" seems to be arbitrary - as long as the > same string is used in amrecover, then it will use the value of > "tpchanger" (AFAIK). > > Hope this helps, > Marc Langlois. > > On Tue, 2004-02-03 at 06:34, R.M. Evers wrote: > > hi amanda-users, > > > > sorry to bring this up again, but i'm also having problems with using > > amrecover with the file: driver.. i have read the recent thread, but i > > cannot get it to work properly. > snip... > > > > then, i tried to recover some directory using amrecover. at the final > > 'extract'-point, i receive the error: > > > > -- > > Extracting files using tape drive daily01 on host bu2. > > Load tape daily02 now > > Continue [?/Y/n/s/t]? > > EOF, check amidxtaped..debug file on bu2. > > amrecover: short block 0 bytes > > UNKNOWN file > > amrecover: Can't read file header > > extract_list - child returned non-zero status: 1 > > -- > > > > the recover works fine when i issue an "amtape DailySet1 label daily02" > > on the backup server, but hey, i'm lazy.. > > > > i've tried different values for the amrecover_changer parameter > > mentioned in the thread last month, but to no avail.. right now, here's > > my config: > > > > -- > > tpchanger "chg-disk" > > changerfile "/etc/amanda/DailySet1/changer" > > tapedev "file:/var/backups/amanda-tapes/DailySet1" > > amrecover_changer "file:/var/backups/amanda-tapes/DailySet1" > > -- > > > > i read the docs, googled like there's no tomorrow, but i can't seem to > > get this right. what am i missing here? > > > > thanks & regards, > > -rodi.
amrecover with file:, again.....
hi amanda-users, sorry to bring this up again, but i'm also having problems with using amrecover with the file: driver.. i have read the recent thread, but i cannot get it to work properly. my backup server is a debian/stable machine. to install amanda-2.4.4p1, i retrieved the source package form debian/unstable. next, i patched the source with the chg-disk patch. next, i created the debian-packages and installed them. so in a way, i created my own patched amanda-backport. i configured the whole thing, and tested 3 dumps with one client server. all went fine :-) then, i tried to recover some directory using amrecover. at the final 'extract'-point, i receive the error: -- Extracting files using tape drive daily01 on host bu2. Load tape daily02 now Continue [?/Y/n/s/t]? EOF, check amidxtaped..debug file on bu2. amrecover: short block 0 bytes UNKNOWN file amrecover: Can't read file header extract_list - child returned non-zero status: 1 -- the recover works fine when i issue an "amtape DailySet1 label daily02" on the backup server, but hey, i'm lazy.. i've tried different values for the amrecover_changer parameter mentioned in the thread last month, but to no avail.. right now, here's my config: -- tpchanger "chg-disk" changerfile "/etc/amanda/DailySet1/changer" tapedev "file:/var/backups/amanda-tapes/DailySet1" amrecover_changer "file:/var/backups/amanda-tapes/DailySet1" -- i read the docs, googled like there's no tomorrow, but i can't seem to get this right. what am i missing here? thanks & regards, -rodi.
Re: 3ware 8506-4 raid1 hotswap
hi christopher, thanks for your reply. what do you mean with 'half a mirror'? i'm planning to use 2 disks, fully mirrored. the first time i pull one out, and put in a fresh disk i expect things to be ok. but what do you think will happen when i put in a disk that has been used before as a full mirror? or is this exactly the situation you're talking about? :-) if i have to clean a disk everytime i put it back in i'm gonna have a problem, since this server will be my only SATA machine... regards, -rodi On Fri, 2004-01-23 at 09:12, Christopher Odenbach wrote: > Pulling one out is ok. We had some problems when inserting a disk that > was used as half a mirror before. This actually killed the machine. The > disk had to be cleaned before reinserting. > > If you experience anything else please let me know. :-) > > Greetings, > > Christopher
3ware 8506-4 raid1 hotswap
hi all, next week, i will be configuring a backup server running Debian, equipped with a 3ware 8506-4 ide-raid controller. the idea is to run a raid1 configuration, allowing for how-swapping of drives as backups. amanda will be using the file driver. does anyone have experience with the 3ware controllers? i would like to know how the hot-swap process is done. is it as simple as pulling a disk out, and putting one back in? or are there some configuration steps involved? regards, -rodi
Re: Replacing a tape
great, i'll do that, thanks! regards, -rodi. On Fri, 2003-12-05 at 13:25, Joshua Baker-LePain wrote: > On Fri, 5 Dec 2003 at 11:37am, R.M. Evers wrote > > > today, i think one of the tapes in my daily backup broke down. it gave > > errors of a bad filedescriptor when i did an amflush. using the next > > tape, things went fine. so i want to replace the tape the next time > > amanda needs it. how do i do this without breaking the order in which > > amanda asks for tapes? > > The day before that tape is next to be used, do this: > > amrmtape $CONFIG $TAPE_LABEL > (and, with a new tape) > amlabel $CONFIT $TAPE_LABEL
Replacing a tape
hi everyone, today, i think one of the tapes in my daily backup broke down. it gave errors of a bad filedescriptor when i did an amflush. using the next tape, things went fine. so i want to replace the tape the next time amanda needs it. how do i do this without breaking the order in which amanda asks for tapes? kind regards, -rodi.
Re: Rackable IDE-RAID1 Harddisk Backup System
thank you paul, this seems fairly straightforward. too bad debian/stable contains amanda-server 2.4.2p2 though.. it's been a long time since i actually had to compile something ;-) now to find a suitable server, and off we go.. thanks again, -rodi. On Thu, 2003-11-13 at 14:04, Paul Bijnens wrote: > R.M. Evers wrote: > > > - is amanda easy to configure to dump to harddisk instead of tapes? > > Yes, using a recent version (2.4.3 at least) with the "file" driver. > See in the manpage under section "OUTPUT DRIVERS". > > You could also leave the files in a large holdingdisk, as Zoltan > suggested, but then there is no easy way to remove old dumps in some > automated way, and keeping the indexes etc synchronised with what is > available to restore. (At least I don't know any.) > > With the file driver, together with a chg-multi changer, you can > emulate all your tapes online. > > In amanda.conf add/change these lines: > = > runtapes 1 > tpchanger "chg-multi" # the tape-changer glue script > #tapedev "/dev/rmt/0n" # the no-rewind tape device to be used > #tapedev "file:/space/amanda-bkup/test" > changerfile "/var/opt/amanda/etc/test/chg-multi.conf" > > amrecover_check_label yes > amrecover_changer "chg-multi" > > > In the "chg-multi.conf" file: > > multieject 0 > gravity 0 > needeject 0 > ejectdelay 0 > statefile /var/opt/amanda/archive/changer-status > firstslot 1 > lastslot 4 > slot 1 file:/space/amanda-bkup/test/Test-01 > slot 2 file:/space/amanda-bkup/test/Test-02 > slot 3 file:/space/amanda-bkup/test/Test-03 > slot 4 file:/space/amanda-bkup/test/Test-04 > = > > Each "slot" in the config file is a directory, which contains > a "data" subdirectory. You have to create these manually. > You can have as many slots (directories) as you like. > The maximum size of each backup run is still limited by > what is specified in the tapetype. You can make up a tapetype > with a suitable capacity. > You still have to "amlabel" each virtual tape, but buying more > tapes is as simple as creating a new directory, with subdir "data" > and adding a new slot in your "chg-multi" configuration. > All the tapes are constatntly online; no need to "insert the > correct tape" etc. > > The filesystem where your backups are stored need to be > able to contain large files (> 4 GByte), unless you can limit > all your backup images to some safe upper bound. (Any recent > Linux will do fine.) > > You still need a holdingdisk, otherwise amanda cannot run multiple dumps > in parallel.
Rackable IDE-RAID1 Harddisk Backup System
hi everyone, i'm new to this list, so please be kind if i'm asking something that has been asked for a thousand times already. but this list seems to be the best place to ask :-) we are running a small ISP, and i've been using amanda for backups for quite a while now and never ran into trouble. it really is a fantastic backup system. backup media has been dds tapes. not too far from now, we are going to move all our servers to a datacenter, which made me decide to buy a brand new rackable hot-swappable ide-raid1 system to do the backups. debian is my OS of choice. my questions are directed to those who have any experience with this kind of setup. so, it boils down to this: - is amanda easy to configure to dump to harddisk instead of tapes? - is there some kind of 'ready-to-go' rackable hot-swappable ide-raid server i can buy (ide-raid compatible with linux, of course..)? thanks in advance. kind regards, rodi.
