2 tapes backup...
Title: 2 tapes backup... My backup have to use 2 different tapes to do full dumps. I read the amanda's docs, but i can't understand how can i tell to amanda to use 2 tapes without tape changer (i use an lto drive so i have to change tape manually at the morning when i arrive to work, tha backup start late at night). I tried with chg-manual, but i can't switch tape when it is ran by cron. Can anyone help me? I am completely lost… Thanks in advance. Giovanni
Re: 2 tapes backup...
On Tuesday 08 November 2005 03:11, Montagni, Giovanni wrote: >My backup have to use 2 different tapes to do full dumps. >I read the amanda's docs, but i can't understand how can i tell to >amanda to use 2 tapes without tape changer (i use an lto drive so i have >to change tape manually at the morning when i arrive to work, tha backup >start late at night). >I tried with chg-manual, but i can't switch tape when it is ran by cron. Why not? The change tape message should be waiting for you in the morning, and when you've changed the tape, amanda should be able to resume, having completed the backup to the holding disk in the meantime. So all it needs to do is flush the remainder. Or, maybe you have to issue the amflush command when you've changed the tape in the morning. I'm not 100% sure, never having used chg-manual personally. You need enough holding disk for that of course. Not having enough holding disk seriously limits amanda's abilities. >Can anyone help me? I am completely lost... > >Thanks in advance. > >Giovanni -- Cheers, Gene "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) 99.35% setiathome rank, not too shabby for a WV hillbilly Free OpenDocument reader/writer/converter download: http://www.openoffice.org Yahoo.com and AOL/TW attorneys please note, additions to the above message by Gene Heskett are: Copyright 2005 by Maurice Eugene Heskett, all rights reserved.
R: 2 tapes backup...
I don't use the holding disk, because i don't have enough free space to store data twice (the data itself and backup). Without holding disk i can't use two tapes? I hope i can... -Messaggio originale- Da: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Per conto di Gene Heskett Inviato: martedì 8 novembre 2005 14.23 A: amanda-users@amanda.org Oggetto: Re: 2 tapes backup... On Tuesday 08 November 2005 03:11, Montagni, Giovanni wrote: >My backup have to use 2 different tapes to do full dumps. >I read the amanda's docs, but i can't understand how can i tell to >amanda to use 2 tapes without tape changer (i use an lto drive so i >have to change tape manually at the morning when i arrive to work, tha >backup start late at night). I tried with chg-manual, but i can't >switch tape when it is ran by cron. Why not? The change tape message should be waiting for you in the morning, and when you've changed the tape, amanda should be able to resume, having completed the backup to the holding disk in the meantime. So all it needs to do is flush the remainder. Or, maybe you have to issue the amflush command when you've changed the tape in the morning. I'm not 100% sure, never having used chg-manual personally. You need enough holding disk for that of course. Not having enough holding disk seriously limits amanda's abilities. >Can anyone help me? I am completely lost... > >Thanks in advance. > >Giovanni -- Cheers, Gene "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) 99.35% setiathome rank, not too shabby for a WV hillbilly Free OpenDocument reader/writer/converter download: http://www.openoffice.org Yahoo.com and AOL/TW attorneys please note, additions to the above message by Gene Heskett are: Copyright 2005 by Maurice Eugene Heskett, all rights reserved.
Re: R: 2 tapes backup...
On Tuesday 08 November 2005 08:54, Montagni, Giovanni wrote: >I don't use the holding disk, because i don't have enough free space to > store data twice (the data itself and backup). > >Without holding disk i can't use two tapes? I hope i can... No, because when amanda is writing direct to tape, and a filesystem dump being written fails due to hitting EOT, then to retry it, the dump itself would have to be restarted from scratch. Amdump doesn't do that. With big disks now a commodity item, expanding the holding disk is a trivial cost task. If your data is valuable, that justify's the cost of another disk. Thats your best argument when standing in the CEO's office asking for a check to go buy another BIG disk. The holding disk will be empty at the end of a successful backup run, so its only temporary usage. And, if you have more than one partition with some free space, they can be defined in your amanda.conf and amanda will use them in rotation as it works. Just be sure to set enough reserved in each one that a full partition never occurs, say 5 or 10%, by setting a figure like -200 mb in that description. That way amanda will use all but the last 200 megs of each area defined. [...] -- Cheers, Gene "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) 99.35% setiathome rank, not too shabby for a WV hillbilly Free OpenDocument reader/writer/converter download: http://www.openoffice.org Yahoo.com and AOL/TW attorneys please note, additions to the above message by Gene Heskett are: Copyright 2005 by Maurice Eugene Heskett, all rights reserved.
Re: R: 2 tapes backup...
On Tue, Nov 08, 2005 at 02:54:20PM +0100, Montagni, Giovanni wrote: >> >> -Messaggio originale- >> Da: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Per conto di Gene Heskett >> Inviato: martedì 8 novembre 2005 14.23 >> A: amanda-users@amanda.org >> Oggetto: Re: 2 tapes backup... >> >> >> On Tuesday 08 November 2005 03:11, Montagni, Giovanni wrote: >> >My backup have to use 2 different tapes to do full dumps. >> >I read the amanda's docs, but i can't understand how can i tell to >> >amanda to use 2 tapes without tape changer (i use an lto drive so i >> >have to change tape manually at the morning when i arrive to work, tha >> >backup start late at night). I tried with chg-manual, but i can't >> >switch tape when it is ran by cron. >> >> Why not? The change tape message should be waiting for you in the morning, >> and when you've changed the tape, amanda should be able to resume, having >> completed the backup to the holding disk in the meantime. So all it >> needs to do is flush the remainder. Or, maybe you have to issue the >> amflush command when you've changed the tape in the morning. I'm not >> 100% sure, never having used chg-manual personally. >> >> You need enough holding disk for that of course. >> Not having enough holding disk seriously limits amanda's abilities. >> > I don't use the holding disk, because i don't have enough free space >> to store data twice (the data itself and backup). > The holding disk only stores untaped data. Not all the data will be on the holding disk at the same time. And if you are using software compression the dumped data will be smaller than the original. While it would be nice to have sufficient free space as your total data, the holding disk need not be that large. >From a cost perspective, your lto drive cost multi-thousand dollars if I read the market correctly. And tapes cost $30 or more each. It would be only a few percentage points of your total backup investment to add an inexpensive disk and dedicate it to holding disk space. The benefits would be improved total time of backup due to parallelism of multiple, simultaneous dumps and completion of the dump before you arrive the next morning. As it stands now, only one DLE (Disk List Entry) can dump at a time. And, even with a second tape, the dump will continue the next morning while people are working. However, you don't need a holding disk to use chg-manual. But there is little to go on from your description of your attempts. What changes did you make to amanda.conf to "try with" chg-manual? Was runtapes one of them? What was your observation that it did not work? Did you read the comments in chg-manual? Hint, always a good idea with script as they may point out some parameters the need to be set. As I recall, by default chg-manual will notify you on your terminal to change the tape. If this is run from cron there will be no terminal. But there is, in the script, a way of changing from terminal messages to email messages. With a holding disk, the next morning when you change the tape, amanda would just dump the untaped archives to the second tape. But without the HD, amanda will resume its dumping, compressing, indexing, network traffic, and taping. All during business hours. In addition to the impact that may make on your users, it also means amanda is backing up file systems while they are actively in use and possibly changing. Never a good idea. No, you don't need a holding disk, but get one! jl -- Jon H. LaBadie [EMAIL PROTECTED] JG Computing 4455 Province Line Road(609) 252-0159 Princeton, NJ 08540-4322 (609) 683-7220 (fax)
R: R: 2 tapes backup...
> -Messaggio originale- > Da: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] Per conto di Jon LaBadie > Inviato: martedì 8 novembre 2005 16.24 > A: amanda-users@amanda.org > Oggetto: Re: R: 2 tapes backup... > > > On Tue, Nov 08, 2005 at 02:54:20PM +0100, Montagni, Giovanni wrote: > >> > >> -Messaggio originale- > >> Da: [EMAIL PROTECTED] > >> [mailto:[EMAIL PROTECTED] Per conto di Gene Heskett > >> Inviato: martedì 8 novembre 2005 14.23 > >> A: amanda-users@amanda.org > >> Oggetto: Re: 2 tapes backup... > >> > >> > >> On Tuesday 08 November 2005 03:11, Montagni, Giovanni wrote: > >> >My backup have to use 2 different tapes to do full dumps. > >> >I read the amanda's docs, but i can't understand how can i tell to > >> >amanda to use 2 tapes without tape changer (i use an lto > drive so i > >> >have to change tape manually at the morning when i arrive > to work, tha > >> >backup start late at night). I tried with chg-manual, but i can't > >> >switch tape when it is ran by cron. > >> > >> Why not? The change tape message should be waiting for you in the > >> morning, and when you've changed the tape, amanda should > be able to > >> resume, having completed the backup to the holding disk in the > >> meantime. So all it needs to do is flush the remainder. > Or, maybe > >> you have to issue the amflush command when you've changed > the tape in > >> the morning. I'm not 100% sure, never having used chg-manual > >> personally. > >> > >> You need enough holding disk for that of course. > >> Not having enough holding disk seriously limits amanda's abilities. > >> > > I don't use the holding disk, because i don't have enough free space > >> to store data twice (the data itself and backup). > > > > The holding disk only stores untaped data. Not all the data > will be on the holding disk at the same time. And if you are > using software compression the dumped data will be smaller > than the original. While it would be nice to have sufficient > free space as your total data, the holding disk need not be > that large. > > From a cost perspective, your lto drive cost multi-thousand > dollars if I read the market correctly. And tapes cost $30 > or more each. It would be only a few percentage points of > your total backup investment to add an inexpensive disk and > dedicate it to holding disk space. > I know it very well, but i can't buy anything. It's another person who decides what to buy... What can i do is ask to him. > The benefits would be improved total time of backup due to > parallelism of multiple, simultaneous dumps and completion of > the dump before you arrive the next morning. As it stands > now, only one DLE (Disk List > Entry) can dump at a time. And, even with a second tape, the > dump will continue the next morning while people are working. > > However, you don't need a holding disk to use chg-manual. > But there is little to go on from your description of your > attempts. What changes did you make to amanda.conf to "try > with" chg-manual? Was runtapes one of them? What was your > observation that it did not work? Did you read the comments > in chg-manual? Hint, always a good idea with script as they > may point out some parameters the need to be set. As I > recall, by default chg-manual will notify you on your > terminal to change the tape. If this is run from cron there > will be no terminal. But there is, in the script, a way of > changing from terminal messages to email messages. I had thought that chg-manual script was provided "as is", so there no way to redirect the notify. Looking into it as you suggest i found that i can change many parameters, as the number of slot, the notify... > > With a holding disk, the next morning when you change the > tape, amanda would just dump the untaped archives to the > second tape. But without the HD, amanda will resume its > dumping, compressing, indexing, network traffic, and taping. > All during business hours. In addition to the impact that may > make on your users, it also means amanda is backing up file > systems while they are actively in use and possibly changing. > Never a good idea. > I run to buy a holding disk! > No, you don't need a holding disk, but get one! > > jl > -- > Jon H. LaBadie [EMAIL PROTECTED] > JG Computing > 4455 Province Line Road(609) 252-0159 > Princeton, NJ 08540-4322 (609) 683-7220 (fax) > Thanks a lot to your help. I understand many many things... Giovanni
"Strange" error message from Amanda.
This time I've got a really strange error message from an amanda back- up: FAILURE AND STRANGE DUMP SUMMARY: epo.dk /var lev 0 STRANGE .. and further down the list: FAILED AND STRANGE DUMP DETAILS: /-- epo.dk /var lev 0 STRANGE sendbackup: start [epo.dk:/var level 0] sendbackup: info BACKUP=/bin/tar sendbackup: info RECOVER_CMD=/usr/bin/gzip -dc |/bin/tar -f... - sendbackup: info COMPRESS_SUFFIX=.gz sendbackup: info end | gtar: ./cache/samba/winbindd_privileged/pipe: socket ignored ? gtar: ./lib/slocate/slocate.db: file changed as we read it | Total bytes written: 545894400 (521MiB, 3.1MiB/s) sendbackup: size 533100 sendbackup: end \ The weird messages do not make any sense what-so-ever. Would someone be kind enough to explain to me what went wrong and whether I have a dependable back-up? Thanks in advance. -- Regards, Erik P. Olsen
Re: "Strange" error message from Amanda.
Erik P. Olsen wrote: | gtar: ./cache/samba/winbindd_privileged/pipe: socket ignored ? gtar: ./lib/slocate/slocate.db: file changed as we read it | Total bytes written: 545894400 (521MiB, 3.1MiB/s) sendbackup: size 533100 sendbackup: end \ The weird messages do not make any sense what-so-ever. Would someone be kind enough to explain to me what went wrong and whether I have a dependable back-up? No need to worry, Erik, just two files that GNUtar could not read, the first a socket-file, the other one changed while dumping. If you want to get rid of the first msg, exclude the whole samba-cache-dir (you don't need to dump it). For the slocate.db: this file could also be generated from scratch, so maybe you want to simply exclude it. OTOH you could shift your cronjob to make sure that this file doesn't get update while amdump runs. Stefan.
Re: "Strange" error message from Amanda.
Erik P. Olsen wrote: > This time I've got a really strange error message from an amanda back- > up: > > FAILURE AND STRANGE DUMP SUMMARY: > epo.dk /var lev 0 STRANGE STRANGE means Amanda received unexpected output from the underlying backup program (tar or dump). Whether or not it is something to be concerned about depends on what it is. See below. > > .. and further down the list: > > FAILED AND STRANGE DUMP DETAILS: > > /-- epo.dk /var lev 0 STRANGE > sendbackup: start [epo.dk:/var level 0] > sendbackup: info BACKUP=/bin/tar > sendbackup: info RECOVER_CMD=/usr/bin/gzip -dc |/bin/tar -f... - > sendbackup: info COMPRESS_SUFFIX=.gz > sendbackup: info end > | gtar: ./cache/samba/winbindd_privileged/pipe: socket ignored Tar is warning you that it came across a special file type (a socket) and didn't back it up. You can safely ignore these messages. > ? gtar: ./lib/slocate/slocate.db: file changed as we read it This means that a file was changing while it was being backed up, so that your backup of that particular file may or may not be valid. This is a hazard of backing up an active filesystem. In this case, I wouldn't worry about it, since it is just the database used by the locate command and can be recreated manually anyway if needed. Other files that might cause the same message are log files, mail spools, and various lock files. You need to look at them and decide if they can safely be ignored, or if is OK to occasionally miss certain files, or if it is a real problem that would force you to reschedule your backups or find other ways to grab the files in question. > | Total bytes written: 545894400 (521MiB, 3.1MiB/s) > sendbackup: size 533100 > sendbackup: end > \ > > The weird messages do not make any sense what-so-ever. Would someone be > kind enough to explain to me what went wrong and whether I have a > dependable back-up? In this example your backups are fine, although sometimes it's good to practice a restore just to make sure the backups are good (and include everything you need). Try a recover into a scratch location first so you don't accidentally trash a server. Frank > > Thanks in advance. -- Frank Smith [EMAIL PROTECTED] Sr. Systems Administrator Voice: 512-374-4673 Hoover's Online Fax: 512-374-4501
Re: "Strange" error message from Amanda.
Frank Smith wrote: Erik P. Olsen wrote: This time I've got a really strange error message from an amanda back- up: FAILURE AND STRANGE DUMP SUMMARY: epo.dk /var lev 0 STRANGE STRANGE means Amanda received unexpected output from the underlying backup program (tar or dump). Whether or not it is something to be concerned about depends on what it is. See below. .. and further down the list: FAILED AND STRANGE DUMP DETAILS: /-- epo.dk /var lev 0 STRANGE sendbackup: start [epo.dk:/var level 0] sendbackup: info BACKUP=/bin/tar sendbackup: info RECOVER_CMD=/usr/bin/gzip -dc |/bin/tar -f... - sendbackup: info COMPRESS_SUFFIX=.gz sendbackup: info end | gtar: ./cache/samba/winbindd_privileged/pipe: socket ignored Tar is warning you that it came across a special file type (a socket) and didn't back it up. You can safely ignore these messages. Not 100% true actually. Gnutar produces some output on stderr which amanda classifies as "normal" (can safely be ignored), "error" (needs to be flagged as error to the user, and implies that this dump is not correct and will not be put on tape if dumping to holdingdisk). And a third category "strange": all the other messages not falling in category "error" or "normal". When amanda detects strange or failed "lines", it shows them all, including the normal ones. The normal ones are preceded by a pipe symbol as above. The "error" and "strange" lines are preceded by a question mark. That means that the above line is not seen as an error. And would that be the only output of tar, amanda would not have bothered you with those details. It is only there to provide some context of the next line (if the error lines need to be interpreted in context, as e.g. samba error message often do, this helps a lot.). ? gtar: ./lib/slocate/slocate.db: file changed as we read it This is the "strange" line. It is not an error. It is up to you to decide if it is bad or not. Frank can explain that better than me: This means that a file was changing while it was being backed up, so that your backup of that particular file may or may not be valid. This is a hazard of backing up an active filesystem. In this case, I wouldn't worry about it, since it is just the database used by the locate command and can be recreated manually anyway if needed. Other files that might cause the same message are log files, mail spools, and various lock files. You need to look at them and decide if they can safely be ignored, or if is OK to occasionally miss certain files, or if it is a real problem that would force you to reschedule your backups or find other ways to grab the files in question. | Total bytes written: 545894400 (521MiB, 3.1MiB/s) sendbackup: size 533100 sendbackup: end \ Notice again the "normal" output line from gnutar: the size line, flagged by amanda with a preceding pipe. -- Paul Bijnens, XplanationTel +32 16 397.511 Technologielaan 21 bus 2, B-3001 Leuven, BELGIUMFax +32 16 397.512 http://www.xplanation.com/ email: [EMAIL PROTECTED] *** * I think I've got the hang of it now: exit, ^D, ^C, ^\, ^Z, ^Q, F6, * * quit, ZZ, :q, :q!, M-Z, ^X^C, logoff, logout, close, bye, /bye, * * stop, end, F3, ~., ^]c, +++ ATH, disconnect, halt, abort, hangup, * * PF4, F20, ^X^X, :D::D, KJOB, F14-f-e, F8-e, kill -1 $$, shutdown, * * kill -9 1, Alt-F4, Ctrl-Alt-Del, AltGr-NumLock, Stop-A, ...* * ... "Are you sure?" ... YES ... Phew ... I'm out * ***
Re: "Strange" error message from Amanda.
On Tue, 2005-11-08 at 22:26 +0100, Paul Bijnens wrote: > >>? gtar: ./lib/slocate/slocate.db: file changed as we read it > > > This is the "strange" line. It is not an error. It is up to you > to decide if it is bad or not. Thanks for shedding light in the darkness. -- Regards, Erik P. Olsen
Re: "Strange" error message from Amanda.
On Tue, Nov 08, 2005 at 10:26:07PM +0100, Paul Bijnens wrote: > > >>| gtar: ./cache/samba/winbindd_privileged/pipe: socket ignored > > > > > >Tar is warning you that it came across a special file type (a socket) > >and didn't back it up. You can safely ignore these messages. > > > Not 100% true actually. Gnutar produces some output on stderr which > amanda classifies as "normal" (can safely be ignored), "error" (needs > to be flagged as error to the user, and implies that this dump is not > correct and will not be put on tape if dumping to holdingdisk). > And a third category "strange": all the other messages not falling in > category "error" or "normal". > > When amanda detects strange or failed "lines", it shows them all, > including the normal ones. The normal ones are preceded by a pipe > symbol as above. The "error" and "strange" lines are preceded by a > question mark. > > That means that the above line is not seen as an error. And would > that be the only output of tar, amanda would not have bothered you > with those details. Darn, that's the second thing this old dog learned today. Thanks for the clarification Paul. -- Jon H. LaBadie [EMAIL PROTECTED] JG Computing 4455 Province Line Road(609) 252-0159 Princeton, NJ 08540-4322 (609) 683-7220 (fax)
Re: "Strange" error message from Amanda.
Paul Bijnens wrote: > Not 100% true actually. Gnutar produces some output on stderr which > amanda classifies as "normal" (can safely be ignored), "error" (needs > to be flagged as error to the user, and implies that this dump is not > correct and will not be put on tape if dumping to holdingdisk). > And a third category "strange": all the other messages not falling in > category "error" or "normal". > > When amanda detects strange or failed "lines", it shows them all, > including the normal ones. The normal ones are preceded by a pipe > symbol as above. The "error" and "strange" lines are preceded by a > question mark. > > That means that the above line is not seen as an error. And would > that be the only output of tar, amanda would not have bothered you > with those details. > It is only there to provide some context of the next line (if the > error lines need to be interpreted in context, as e.g. samba error > message often do, this helps a lot.). > Well, I stand corrected (and learned something in the process). I was going to make a comment about this behavior not being adequately documented, but it looks like it is, although in the dumperapi.txt file, which I had never looked at. There is a mention of the '?' in using.txt but it doesn't explain '|' (normal) vs, '?' (strange) vs. '&' (error) as the dumperapi.txt does. If the explanation hasn't been included in the general use docs yet (I'm looking at a 2.4.5 source tree), it should be. Thanks for the enlightenment, Frank -- Frank Smith [EMAIL PROTECTED] Sr. Systems Administrator Voice: 512-374-4673 Hoover's Online Fax: 512-374-4501
