Re: [Bacula-users] bacula prunes files too early
Really UTF8. However, tests I did was with simple filenames (1,2,3,4 etc)... Whether or no, I also converted database and will supervise. Hello, Is your PostgreSQL database encoding is SQL_ASCII ? Do you see error in postgres log like that : ERROR: invalid byte sequence for encoding UTF8: If so, look at there : http://article.gmane.org/gmane.comp.sysutils.backup.bacula.devel/12074/ I think it was my problem. I've just switch my database to SQL_ASCII and I'm waiting for the next full backup to occur to see if it works. Regards. -- Vladimir Vassiliev v...@edu.yar.ru -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] How can I prevent this?
Arno Lehmann wrote: There are some ideas floating around in the list archives - the most interesting one, to me, is to use a run before job script that, using bconsole, checks which jobs are currently running and aborts with an error code if its own name is in the list. Then, you can wait for the next major version of Bacula, where there will be configuration directives to modify the scheduling and running behaviour in thss case. Arno Ok, I can wait. Are the new directives documented yet? -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] How can I prevent this?
-Original Message- From: Brian Debelius [mailto:bdebel...@intelesyscorp.com] Sent: Friday, February 06, 2009 7:22 AM To: Arno Lehmann Cc: bacula-users Subject: Re: [Bacula-users] How can I prevent this? Arno Lehmann wrote: There are some ideas floating around in the list archives - the most interesting one, to me, is to use a run before job script that, using bconsole, checks which jobs are currently running and aborts with an error code if its own name is in the list. Then, you can wait for the next major version of Bacula, where there will be configuration directives to modify the scheduling and running behaviour in thss case. Arno Ok, I can wait. Are the new directives documented yet? Yes, please see http://www.bacula.org/manuals/en/concepts/concepts/New_Features.html#SEC TION0057 -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] How can I prevent this?
Robert LeBlanc wrote: Yes, please see http://www.bacula.org/manuals/en/concepts/concepts/New_Features.html#SEC TION0057000 I don't think this will handle this case. Lets say I have the following nightly backups Priority 10 BackupA Priority 11 BackupB Priority 12 BackupC That are schedule for tonight Priority 10 09-FEB-09 00:01BackupA Priority 11 09-FEB-09 00:01BackupB Priority 12 09-FEB-09 00:01BackupC And BackupA starts running but for some reason gets distracted. Priority 10 09-FEB-09 00:01BackupA *procrastinating* Priority 11 09-FEB-09 00:01BackupB *waiting for a higher priority job* Priority 12 09-FEB-09 00:01BackupC *waiting for a higher priority job* Now time has past and its February 10. (I think this is what happened, before) Priority 10 09-FEB-09 00:01BackupA *procrastinating* Priority 10 10-FEB-09 00:01BackupA Priority 11 09-FEB-09 00:01BackupB Priority 11 10-FEB-09 00:01BackupB Priority 12 09-FEB-09 00:01BackupC Priority 12 10-FEB-09 00:01BackupC But I want it to be scheduled like, so the queue is run by date/time, then priority Priority 10 09-FEB-09 00:01BackupA *procrastinating* Priority 11 09-FEB-09 00:01BackupB Priority 12 09-FEB-09 00:01BackupC Priority 10 10-FEB-09 00:01BackupA Priority 11 10-FEB-09 00:01BackupB Priority 12 10-FEB-09 00:01BackupC -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
[Bacula-users] Big Restore Jobs is fast, Small Restore Jobs is slow =(
Hello! After few tests i discovered that my bacula have an strange issue.. if i run an restore for an job around 20GB or more my restore is faster 11mbps, i can restore it in a few minutes If i run an restore small like 10MB, the restore runs about 10kpbs, if more small more slow.. My bacula is 1.38 running on an freebsd. The client in these tests is a Win2003. my list jobs is here: http://rafb.net/p/xdSDX523.html What can be wrong? -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] Using tape autochanger
I think this: Here is the storage related sections of -dir.conf in directory /etc/bacula: Storage { Name = LTO-2 Address = host.domain.com SDPort = 9103 Password = host Device = /dev/st0 Media Type = Ultrium2 } Should be: Storage { Name = LTO-2 Address = host.domain.com SDPort = 9103 Password = host Device = Autochanger Autochanger = yes Media Type = Ultrium2 } Christopher Dick wrote: I will preface this with the usual I am new to Bacula statement before diving into this. I have Googled the errors, and looked at email threads and attempted to get my configs to look like the ones in the documentation and in the various forums and email threads. But nothing seems to work. Here is the issue: When I run btape, I get the following: btape -c /etc/bacula/bacula-sd.conf autochanger Tape block granularity is 1024 bytes. btape: butil.c:278 Could not find device autochanger in config file /etc/bacula/bacula-sd.conf. 05-Feb 16:39 btape JobId 0: Fatal error: butil.c:168 Cannot find device autochanger in config file /etc/bacula/bacula-sd.conf. When I run bconsole, I regularly get this: 3999 Device LTO-2 not found or could not be opened. Here is the storage related sections of -dir.conf in directory /etc/bacula: Storage { Name = LTO-2 Address = host.domain.com SDPort = 9103 Password = host Device = /dev/st0 Media Type = Ultrium2 } Here is the Pool definition in -dir.conf Pool { Name = Tape Pool Type = Backup Volume Retention = 365 days Recycle = yes AutoPrune = yes Storage = LTO-2 } And finally, here are the entries in -sd.conf from /etc/bacula: Autochanger { Name = Autochanger Device = LTO-2 Changer Command = /etc/bacula/mtx-changer %c %o %S %a %d Changer Device = /dev/sg1 } Device { Name = LTO-2 Drive Index = 0 Device Type = Tape Media Type = Ultrium2 Archive Device = /dev/st0 AutomaticMount = yes AlwaysOpen = yes RemovableMedia = yes RandomAccess = no; AutoChanger = yes LabelMedia = yes Spool Directory = /tmp/spool Maximum Spool Size = 7 Random Access = no } dmesg has the following information: scsi 1:0:0:0: Medium Changer HP C7200 1720 PQ: 0 ANSI: 3 scsi 1:0:0:0: Attached scsi generic sg1 type 8 scsi 1:0:0:1: Sequential-Access HP Ultrium 2-SCSI F59W PQ: 0 ANSI: 3 st 1:0:0:1: Attached scsi tape st0 st 1:0:0:1: st0: try direct i/o: yes (alignment 4 B) st 1:0:0:1: Attached scsi generic sg2 type 1 This is running on a dedicated backup server running openSUSE with Bacula v2.4.2, with an HP C7200 Galactica model tape library, upgraded to LTO-2. Currently it only has one drive, but will soon have two. Currently, Bacula is running as root (I will change that once I get this working), but I have chmod'd /dev/st0 /dev/nst0 and /dev/sg2 to 0777 just to be sure. I've tried using /dev/nst0 and /dev/sg2, the latter being what the HP Tape Tools reports as being the actual tape drive and matches dmesg, but nothing seems to go. Hopefully that is enough information. if not, I will be happy to provide anything else needed. Any assistance is greatly appreciated. I am so close to getting this working, I can taste it. Thanks! -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] Big Restore Jobs is fast, Small Restore Jobs is slow =(
On Fri, Feb 6, 2009 at 11:26 AM, Henrique Machado henri@gmail.com wrote: Sorry I forgot to say that i'm using Hard Disk's to storage.. Is easy to upgrade? I do not use freebsd but others on this list do. Dan Langille should be of some help. John -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] atoi and volume name/media name ambiguity (Was: Re: Bug / RFE)
Bailey, Scott wrote: I second the notion that numeric volume labels are in widespread use. Maybe those barcode manufacturers gave special discounts to people who ordered ranges without alphabetic characters in them. :-) My suggested kludge actually would be the opposite of what others have suggested; namely, I would assume first that user input actually is a literal label and consider it as a potential MediaId only if no match is found. First, note that Bacula does not use the atoi() function to convert the string, but rather its own str_to_int64() function. I would imagine this is in order to handle the input of 64-bit values on 32-bit machines. Since it is possible to have a numeric VolumeLabel, it simply is not possible to distinguish between MediaId and VolumeLabel when given a string. For example, VolumeLabel 1234 and MediaId 1234 are indistinguishable given the string 1234. When using strictly numeric VolumeLabels, it is very likely that a volume with VolumeLabel 1234 and a different volume with MediaId 1234 both exist at the same time. One way to deal with this would be to allow quoting for a VolumeLabel entry. Another way is to force the user to include a leading '+' when entering a MediaId and considering strings that do not begin with a '+' to be a VolumeLabel. Neither way is foolproof when using strictly numeric volume labels. Both ways are more cumbersome, break existing scripts (or anything that invokes bconsole to send commands to Dir), and likely are not needed by the majority of people. Forcing a leading '+' is a simple one line change. I would in no way consider this a bug fix, but for those who want to consider a numeric string a VolumeLabel unless a leading '+' is entered, a patch against the ua_select.c in trunk (as of today) is below. --- bacula/src/dird/ua_select.c 2009-02-06 09:19:28.0 -0500 +++ ua_select.c 2009-02-06 11:57:22.0 -0500 @@ -573,7 +573,7 @@ if (!get_cmd(ua, _(Enter MediaId or Volume name: ))) { return 0; } - if (is_a_number(ua-cmd)) { + if (ua-cmd[0] == '+' is_a_number(ua-cmd)) { mr-MediaId = str_to_int64(ua-cmd); } else { bstrncpy(mr-VolumeName, ua-cmd, sizeof(mr-VolumeName)); Cheers, Josh Fisher Further, although you can't do much for existing installations, we could reduce ambiguity by starting with MediaId 100 (rather than 1) -- so that valid MediaIds cannot be encoded in a six-character ANSI label. That's not a total cure, I've not seen longer labels in our environments. Actually, suitably desperate individuals probably could perform some SQL magic on their database now to renumber MediaIds to a less dangerous range, and avoid the risk of bacula doing something you did not intend. That's still no fix for mangling your input and refusing to do what you really wanted. ;-) Cheers, Scott Bailey scott.bai...@eds.com -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
[Bacula-users] Incomplete backup
Backing up a mail server I realized that the backup is incomplete. I use bacula 2.4.4. Bacula-fd runs as root. My FileSet is like this: FileSet { Name = mail Include { Options { signature = MD5 compression = GZIP } File = /var/zimbra } } Before backup I checked the directory /var/zimbra/store/0/10/msg/0/ and it contained 2043 files. After the backup, in the bacula console I see that the backup contains only 4 files: cwd is: /var/zimbra/store/0/10/msg/0/ $ ls 3029-8857.msg 3080-9300.msg 3081-9301.msg 3083-9326.msg What can be the reason for this? Thank you. -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] Incomplete backup
I read this link. http://article.gmane.org/gmane.comp.sysutils.backup.bacula.devel/12074/ I use postgresql, but my database is SQL_ASCII, so it seems that I don't fit in this bug report. Victor Sterpu wrote: Backing up a mail server I realized that the backup is incomplete. I use bacula 2.4.4. Bacula-fd runs as root. My FileSet is like this: FileSet { Name = mail Include { Options { signature = MD5 compression = GZIP } File = /var/zimbra } } Before backup I checked the directory /var/zimbra/store/0/10/msg/0/ and it contained 2043 files. After the backup, in the bacula console I see that the backup contains only 4 files: cwd is: /var/zimbra/store/0/10/msg/0/ $ ls 3029-8857.msg 3080-9300.msg 3081-9301.msg 3083-9326.msg What can be the reason for this? Thank you. -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] Incomplete backup
I runned bacula in debug mode and found out something strange. bacula-dir -c /etc/bacula/bacula-dir.conf -d99 -f -u bacula bacula.out The command cat bacula.out | grep postgres returns this: 192.168.0.191-dir: postgresql.c:194-0 pg_real_connect done 192.168.0.191-dir: postgresql.c:196-0 db_user=bacula db_name=bacula db_password= 192.168.0.191-dir: postgresql.c:513-0 Result status failed: set standard_conforming_strings=on ... 192.168.0.191-dir: postgresql.c:513-0 Result status failed: DROP TABLE DelCandidates 192.168.0.191-dir: postgresql.c:513-0 Result status failed: DROP INDEX DelInx1 192.168.0.191-dir: postgresql.c:513-0 Result status failed: DROP INDEX DelInx1 The postgres log: 2009-02-06 22:32:37 EET ERROR: parameter standard_conforming_strings cannot be changed 2009-02-06 22:32:51 EET ERROR: parameter standard_conforming_strings cannot be changed 2009-02-06 22:33:38 EET ERROR: table delcandidates does not exist 2009-02-06 22:33:38 EET ERROR: index delinx1 does not exist 2009-02-06 22:33:38 EET ERROR: index delinx1 does not exist Victor Sterpu wrote: Backing up a mail server I realized that the backup is incomplete. I use bacula 2.4.4. Bacula-fd runs as root. My FileSet is like this: FileSet { Name = mail Include { Options { signature = MD5 compression = GZIP } File = /var/zimbra } } Before backup I checked the directory /var/zimbra/store/0/10/msg/0/ and it contained 2043 files. After the backup, in the bacula console I see that the backup contains only 4 files: cwd is: /var/zimbra/store/0/10/msg/0/ $ ls 3029-8857.msg 3080-9300.msg 3081-9301.msg 3083-9326.msg What can be the reason for this? Thank you. -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
[Bacula-users] why can t running multiples jobs ?
hello i ve the following conf : Device status: Autochanger 136T with devices: Drive-0 (/dev/st0) Drive-1 (/dev/st1) Drive-2 (/dev/st2) Device Drive-0 (/dev/st0) is not open. Drive 0 status unknown. Device Drive-1 (/dev/st1) is not open. Drive 1 status unknown. Device Drive-2 (/dev/st2) is mounted with: Volume: LT1008L3 Pool:Default Media type: LTO-3 Slot 39 is loaded in drive 2. Total Bytes=23,531,268,096 Blocks=364,757 Bytes/block=64,512 Positioned at File=36 Block=11,641 3 drives , in an autoloader . And , always ... only one , drive working . not 3 as the same time (with 3 differents jobs ... ) why ? where is the mistake ? regards -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] why can t running multiples jobs ?
On Fri, Feb 6, 2009 at 3:57 PM, gui...@free.fr wrote: hello i ve the following conf : Device status: Autochanger 136T with devices: Drive-0 (/dev/st0) Drive-1 (/dev/st1) Drive-2 (/dev/st2) Device Drive-0 (/dev/st0) is not open. Drive 0 status unknown. Device Drive-1 (/dev/st1) is not open. Drive 1 status unknown. Device Drive-2 (/dev/st2) is mounted with: Volume: LT1008L3 Pool:Default Media type: LTO-3 Slot 39 is loaded in drive 2. Total Bytes=23,531,268,096 Blocks=364,757 Bytes/block=64,512 Positioned at File=36 Block=11,641 3 drives , in an autoloader . And , always ... only one , drive working . not 3 as the same time (with 3 differents jobs ... ) why ? where is the mistake ? regards Probably because your jobs all go to the same pool and bacula already has the pool loaded in one drive so it does not think it needs to find more volumes of the same pool and use them on different jobs in different drives. John -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] Incomplete backup
Victor Sterpu schrieb: Backing up a mail server I realized that the backup is incomplete. I use bacula 2.4.4. Bacula-fd runs as root. My FileSet is like this: FileSet { Name = mail Include { Options { signature = MD5 compression = GZIP } File = /var/zimbra } } Before backup I checked the directory /var/zimbra/store/0/10/msg/0/ and it contained 2043 files. After the backup, in the bacula console I see that the backup contains only 4 files: cwd is: /var/zimbra/store/0/10/msg/0/ $ ls 3029-8857.msg 3080-9300.msg 3081-9301.msg 3083-9326.msg What can be the reason for this? You are running a full backup? Not an incremental? Show us the job output and final job report. Ralf -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] why can t running multiples jobs ?
On Fri, Feb 06, 2009 at 09:57:08PM +0100, gui...@free.fr wrote: hello i ve the following conf : Device status: Autochanger 136T with devices: Drive-0 (/dev/st0) Drive-1 (/dev/st1) Drive-2 (/dev/st2) Device Drive-0 (/dev/st0) is not open. Drive 0 status unknown. Device Drive-1 (/dev/st1) is not open. Drive 1 status unknown. Device Drive-2 (/dev/st2) is mounted with: Volume: LT1008L3 Pool:Default Media type: LTO-3 Slot 39 is loaded in drive 2. Total Bytes=23,531,268,096 Blocks=364,757 Bytes/block=64,512 Positioned at File=36 Block=11,641 3 drives , in an autoloader . And , always ... only one , drive working . not 3 as the same time (with 3 differents jobs ... ) why ? where is the mistake ? regards how do you have concurrency configured? -- michael -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] why can t running multiples jobs ?
On Fri, Feb 06, 2009 at 04:11:11PM -0500, John Drescher wrote: On Fri, Feb 6, 2009 at 3:57 PM, gui...@free.fr wrote: hello i ve the following conf : Device status: Autochanger 136T with devices: Drive-0 (/dev/st0) Drive-1 (/dev/st1) Drive-2 (/dev/st2) Device Drive-0 (/dev/st0) is not open. Drive 0 status unknown. Device Drive-1 (/dev/st1) is not open. Drive 1 status unknown. Device Drive-2 (/dev/st2) is mounted with: Volume: LT1008L3 Pool:Default Media type: LTO-3 Slot 39 is loaded in drive 2. Total Bytes=23,531,268,096 Blocks=364,757 Bytes/block=64,512 Positioned at File=36 Block=11,641 3 drives , in an autoloader . And , always ... only one , drive working . not 3 as the same time (with 3 differents jobs ... ) why ? where is the mistake ? regards Probably because your jobs all go to the same pool and bacula already has the pool loaded in one drive so it does not think it needs to find more volumes of the same pool and use them on different jobs in different drives. then in that case Prefered Mounted Volumes = No will help perhaps? -- michael -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] Incomplete backup
Ralf Gross wrote: Victor Sterpu schrieb: Backing up a mail server I realized that the backup is incomplete. I use bacula 2.4.4. Bacula-fd runs as root. My FileSet is like this: FileSet { Name = mail Include { Options { signature = MD5 compression = GZIP } File = /var/zimbra } } Before backup I checked the directory /var/zimbra/store/0/10/msg/0/ and it contained 2043 files. After the backup, in the bacula console I see that the backup contains only 4 files: cwd is: /var/zimbra/store/0/10/msg/0/ $ ls 3029-8857.msg 3080-9300.msg 3081-9301.msg 3083-9326.msg What can be the reason for this? You are running a full backup? Not an incremental? Show us the job output and final job report. Ralf -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] Incomplete backup
I run a incremental job, but when I try to restore I should see all the files. I did this at the bacula console: restore 5: Select the most recent backup for a client Automatically selected FileSet: mail +---+---+--+-+-+-+ | jobid | level | jobfiles | jobbytes| starttime | volumename | +---+---+--+-+-+-+ |54 | F | 29,052 | 366,387,257 | 2009-02-06 02:18:08 | StorageMail | |55 | I | 327 | 69,960,993 | 2009-02-06 07:23:52 | StorageMail | |60 | I | 944 | 369,334,548 | 2009-02-06 21:48:43 | StorageMail | |63 | I | 86 | 123,432,845 | 2009-02-06 22:07:21 | StorageMail | |64 | I | 198 | 124,352,818 | 2009-02-06 22:32:50 | StorageMail | |66 | I | 222 | 99,816,375 | 2009-02-06 23:16:33 | StorageMail | +---+---+--+-+-+-+ You have selected the following JobIds: 54,55,60,63,64,66 Building directory tree for JobId 54 ... + Building directory tree for JobId 55 ... Building directory tree for JobId 60 ... + Building directory tree for JobId 63 ... Building directory tree for JobId 64 ... Building directory tree for JobId 66 ... 6 Jobs, 29,676 files inserted into the tree. You are now entering file selection mode where you add (mark) and remove (unmark) files to be restored. No files are initially added, unless you used the all keyword on the command line. Enter done to leave this mode. cwd is: / $ cd /var/zimbra/store/0/10/msg/0/ cwd is: /var/zimbra/store/0/10/msg/0/ $ ls 3029-8857.msg 3080-9300.msg 3081-9301.msg 3083-9326.msg ls -al /var/zimbra/store/0/10/msg/0/ | wc returns 2043 16338 136026 and all of them are files(mail messages). The final job report is like this: 06-Feb 23:16 192.168.0.191-dir JobId 66: Start Backup JobId 66, Job=Backup_Mail.2009-02-06_23.16.30.04 06-Feb 23:16 192.168.0.191-dir JobId 66: Using Device FSZimbra 06-Feb 23:16 192.168.0.191-sd JobId 66: Volume StorageMail previously written, moving to end of data. 06-Feb 23:16 192.168.0.191-sd JobId 66: Ready to append to end of Volume StorageMail size=1060249931 06-Feb 23:17 192.168.0.191-sd JobId 66: Job write elapsed time = 00:00:42, Transfer rate = 2.377 M bytes/second 06-Feb 23:17 192.168.0.191-dir JobId 66: Bacula 192.168.0.191-dir 2.4.4 (28Dec08): 06-Feb-2009 23:17:16 Build OS: i686-pc-linux-gnu debian 4.0 JobId: 66 Job:Backup_Mail.2009-02-06_23.16.30.04 Backup Level: Incremental, since=2009-02-06 22:32:50 Client: mail-fd 2.4.4 (28Dec08) i686-pc-linux-gnu,debian,4.0 FileSet:mail 2009-02-05 23:22:00 Pool: PMail (From Job resource) Storage:SMail (From Pool resource) Scheduled time: 06-Feb-2009 23:16:29 Start time: 06-Feb-2009 23:16:33 End time: 06-Feb-2009 23:17:16 Elapsed time: 43 secs Priority: 10 FD Files Written: 222 SD Files Written: 222 FD Bytes Written: 99,816,375 (99.81 MB) SD Bytes Written: 99,847,184 (99.84 MB) Rate: 2321.3 KB/s Software Compression: 67.5 % VSS:no Storage Encryption: no Volume name(s): StorageMail Volume Session Id: 10 Volume Session Time:1233948659 Last Volume Bytes: 1,160,216,405 (1.160 GB) Non-fatal FD errors:0 SD Errors: 0 FD termination status: OK SD termination status: OK Termination:Backup OK 06-Feb 23:17 192.168.0.191-dir JobId 66: Begin pruning Jobs. 06-Feb 23:17 192.168.0.191-dir JobId 66: No Jobs found to prune. 06-Feb 23:17 192.168.0.191-dir JobId 66: Begin pruning Files. 06-Feb 23:17 192.168.0.191-dir JobId 66: No Files found to prune. 06-Feb 23:17 192.168.0.191-dir JobId 66: End auto prune. Ralf Gross wrote: Victor Sterpu schrieb: Backing up a mail server I realized that the backup is incomplete. I use bacula 2.4.4. Bacula-fd runs as root. My FileSet is like this: FileSet { Name = mail Include { Options { signature = MD5 compression = GZIP } File = /var/zimbra } } Before backup I checked the directory /var/zimbra/store/0/10/msg/0/ and it contained 2043 files. After the backup, in the bacula console I see that the backup contains only 4 files: cwd is: /var/zimbra/store/0/10/msg/0/ $ ls 3029-8857.msg 3080-9300.msg 3081-9301.msg 3083-9326.msg What can be the reason for this? You are running a full backup? Not an incremental? Show us the job output and final job report. Ralf -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR,
Re: [Bacula-users] Incomplete backup
I run a incremental job, but when I try to restore I should see all the files. I did this at the bacula console: restore 5: Select the most recent backup for a client Automatically selected FileSet: mail +---+---+--+-+-+-+ | jobid | level | jobfiles | jobbytes| starttime | volumename | +---+---+--+-+-+-+ |54 | F | 29,052 | 366,387,257 | 2009-02-06 02:18:08 | StorageMail | |55 | I | 327 | 69,960,993 | 2009-02-06 07:23:52 | StorageMail | |60 | I | 944 | 369,334,548 | 2009-02-06 21:48:43 | StorageMail | |63 | I | 86 | 123,432,845 | 2009-02-06 22:07:21 | StorageMail | |64 | I | 198 | 124,352,818 | 2009-02-06 22:32:50 | StorageMail | |66 | I | 222 | 99,816,375 | 2009-02-06 23:16:33 | StorageMail | +---+---+--+-+-+-+ You have selected the following JobIds: 54,55,60,63,64,66 Building directory tree for JobId 54 ... + Building directory tree for JobId 55 ... Building directory tree for JobId 60 ... + Building directory tree for JobId 63 ... Building directory tree for JobId 64 ... Building directory tree for JobId 66 ... 6 Jobs, 29,676 files inserted into the tree. You are now entering file selection mode where you add (mark) and remove (unmark) files to be restored. No files are initially added, unless you used the all keyword on the command line. Enter done to leave this mode. cwd is: / $ cd /var/zimbra/store/0/10/msg/0/ cwd is: /var/zimbra/store/0/10/msg/0/ $ ls 3029-8857.msg 3080-9300.msg 3081-9301.msg 3083-9326.msg ls -al /var/zimbra/store/0/10/msg/0/ | wc returns 2043 16338 136026 and all of them are files(mail messages). The final job report is like this: 06-Feb 23:16 192.168.0.191-dir JobId 66: Start Backup JobId 66, Job=Backup_Mail.2009-02-06_23.16.30.04 06-Feb 23:16 192.168.0.191-dir JobId 66: Using Device FSZimbra 06-Feb 23:16 192.168.0.191-sd JobId 66: Volume StorageMail previously written, moving to end of data. 06-Feb 23:16 192.168.0.191-sd JobId 66: Ready to append to end of Volume StorageMail size=1060249931 06-Feb 23:17 192.168.0.191-sd JobId 66: Job write elapsed time = 00:00:42, Transfer rate = 2.377 M bytes/second 06-Feb 23:17 192.168.0.191-dir JobId 66: Bacula 192.168.0.191-dir 2.4.4 (28Dec08): 06-Feb-2009 23:17:16 Build OS: i686-pc-linux-gnu debian 4.0 JobId: 66 Job:Backup_Mail.2009-02-06_23.16.30.04 Backup Level: Incremental, since=2009-02-06 22:32:50 Client: mail-fd 2.4.4 (28Dec08) i686-pc-linux-gnu,debian,4.0 FileSet:mail 2009-02-05 23:22:00 Pool: PMail (From Job resource) Storage:SMail (From Pool resource) Scheduled time: 06-Feb-2009 23:16:29 Start time: 06-Feb-2009 23:16:33 End time: 06-Feb-2009 23:17:16 Elapsed time: 43 secs Priority: 10 FD Files Written: 222 SD Files Written: 222 FD Bytes Written: 99,816,375 (99.81 MB) SD Bytes Written: 99,847,184 (99.84 MB) Rate: 2321.3 KB/s Software Compression: 67.5 % VSS:no Storage Encryption: no Volume name(s): StorageMail Volume Session Id: 10 Volume Session Time:1233948659 Last Volume Bytes: 1,160,216,405 (1.160 GB) Non-fatal FD errors:0 SD Errors: 0 FD termination status: OK SD termination status: OK Termination:Backup OK 06-Feb 23:17 192.168.0.191-dir JobId 66: Begin pruning Jobs. 06-Feb 23:17 192.168.0.191-dir JobId 66: No Jobs found to prune. 06-Feb 23:17 192.168.0.191-dir JobId 66: Begin pruning Files. 06-Feb 23:17 192.168.0.191-dir JobId 66: No Files found to prune. 06-Feb 23:17 192.168.0.191-dir JobId 66: End auto prune. Ralf Gross wrote: Victor Sterpu schrieb: Backing up a mail server I realized that the backup is incomplete. I use bacula 2.4.4. Bacula-fd runs as root. My FileSet is like this: FileSet { Name = mail Include { Options { signature = MD5 compression = GZIP } File = /var/zimbra } } Before backup I checked the directory /var/zimbra/store/0/10/msg/0/ and it contained 2043 files. After the backup, in the bacula console I see that the backup contains only 4 files: cwd is: /var/zimbra/store/0/10/msg/0/ $ ls 3029-8857.msg 3080-9300.msg 3081-9301.msg 3083-9326.msg What can be the reason for this? You are running a full backup? Not an incremental? Show us the job output and final job report. Ralf -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR,
Re: [Bacula-users] Using tape autochanger
- Original Message From: Brian Debelius bdebel...@intelesyscorp.com To: Christopher Dick jcdi...@yahoo.com Cc: bacula-users@lists.sourceforge.net Sent: Friday, February 6, 2009 10:28:51 AM Subject: Re: [Bacula-users] Using tape autochanger I think this: Here is the storage related sections of -dir.conf in directory /etc/bacula: Storage { Name = LTO-2 Address = host.domain.com SDPort = 9103 Password = host Device = /dev/st0 Media Type = Ultrium2 } Should be: Storage { Name = LTO-2 Address = host.domain.com SDPort = 9103 Password = host Device = Autochanger Autochanger = yes Media Type = Ultrium2 } Christopher Dick wrote: I will preface this with the usual I am new to Bacula statement before diving into this. I have Googled the errors, and looked at email threads and attempted to get my configs to look like the ones in the documentation and in the various forums and email threads. But nothing seems to work. Here is the issue: When I run btape, I get the following: btape -c /etc/bacula/bacula-sd.conf autochanger Tape block granularity is 1024 bytes. btape: butil.c:278 Could not find device autochanger in config file /etc/bacula/bacula-sd.conf. 05-Feb 16:39 btape JobId 0: Fatal error: butil.c:168 Cannot find device autochanger in config file /etc/bacula/bacula-sd.conf. When I run bconsole, I regularly get this: 3999 Device LTO-2 not found or could not be opened. Here is the storage related sections of -dir.conf in directory /etc/bacula: Storage { Name = LTO-2 Address = host.domain.com SDPort = 9103 Password = host Device = /dev/st0 Media Type = Ultrium2 } Here is the Pool definition in -dir.conf Pool { Name = Tape Pool Type = Backup Volume Retention = 365 days Recycle = yes AutoPrune = yes Storage = LTO-2 } And finally, here are the entries in -sd.conf from /etc/bacula: Autochanger { Name = Autochanger Device = LTO-2 Changer Command = /etc/bacula/mtx-changer %c %o %S %a %d Changer Device = /dev/sg1 } Device { Name = LTO-2 Drive Index = 0 Device Type = Tape Media Type = Ultrium2 Archive Device = /dev/st0 AutomaticMount = yes AlwaysOpen = yes RemovableMedia = yes RandomAccess = no; AutoChanger = yes LabelMedia = yes Spool Directory = /tmp/spool Maximum Spool Size = 7 Random Access = no } dmesg has the following information: scsi 1:0:0:0: Medium Changer HP C7200 1720 PQ: 0 ANSI: 3 scsi 1:0:0:0: Attached scsi generic sg1 type 8 scsi 1:0:0:1: Sequential-Access HP Ultrium 2-SCSI F59W PQ: 0 ANSI: 3 st 1:0:0:1: Attached scsi tape st0 st 1:0:0:1: st0: try direct i/o: yes (alignment 4 B) st 1:0:0:1: Attached scsi generic sg2 type 1 This is running on a dedicated backup server running openSUSE with Bacula v2.4.2, with an HP C7200 Galactica model tape library, upgraded to LTO-2. Currently it only has one drive, but will soon have two. Currently, Bacula is running as root (I will change that once I get this working), but I have chmod'd /dev/st0 /dev/nst0 and /dev/sg2 to 0777 just to be sure. I've tried using /dev/nst0 and /dev/sg2, the latter being what the HP Tape Tools reports as being the actual tape drive and matches dmesg, but nothing seems to go. Hopefully that is enough information. if not, I will be happy to provide anything else needed. Any assistance is greatly appreciated. I am so close to getting this working, I can taste it. Thanks! You were right. That was definitely the issue. I misunderstood the documentation in regard to this. Thanks for catching it! My backups are now working just fine, for the most part. I will submit other emails as questions arise, so as not to hijack this thread and take it away from the simple subject and solution. Thanks a bunch! Chris -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
[Bacula-users] Overwriting previous fulls and other miscellaneous tweaks
Now that I have my tape autochanger and such working as desired, though it may take a little tweaking, I am overall pretty stoked about getting this working. Now to start getting into the gritty details of tweaking for my environment and limited capacities. My first question is in regard to the monthly fulls. I have an issue with tape population in my library. I have only 20 slots for LTO-2 tapes, and my first full backup of my primary server has already used six of those. The data doesn't compress very well. Is there a way to inform Bacula to overwrite only that system's previous full when it goes to do a new one? We have TSM at my place of employment, and it does the whole one time full, incremental forever thing, which would reduce dramatically the number of tapes required. However, I am pretty limited on my available resources as far as tapes are concerned, and so I don't have the ability to maintain a stack of 60 unused tapes waiting to check out full ones and refill with empties. Two months will fill all my available tapes. I am basically needing to do a poor man's de-dupe, really, so that there is only one copy of a server's data in the library at any given time, even three, four or five years from now. Is that possible? Also, I am using spooling to maximize performance on the drives. I have the max bytes set to about 90% of my spool disks' capacity. However, I didn't see anywhere if there was a low bytes setting, so that if the pool fills, it will spool off to tape all the way down to the minimum instead of just until the spool says I'm not at my max capacity anymore. Or does it do that and I am just not setting the max size right and have it a decimal place too small? Any help is appreciated. And this Bacula thing is really a god-send. Thanks! Chris -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] Overwriting previous fulls and other miscellaneous tweaks
On Fri, Feb 6, 2009 at 7:29 PM, Christopher Dick jcdi...@yahoo.com wrote: Now that I have my tape autochanger and such working as desired, though it may take a little tweaking, I am overall pretty stoked about getting this working. Now to start getting into the gritty details of tweaking for my environment and limited capacities. My first question is in regard to the monthly fulls. I have an issue with tape population in my library. I have only 20 slots for LTO-2 tapes, and my first full backup of my primary server has already used six of those. The data doesn't compress very well. Is there a way to inform Bacula to overwrite only that system's previous full when it goes to do a new one? We have TSM at my place of employment, and it does the whole one time full, incremental forever thing, which would reduce dramatically the number of tapes required. However, I am pretty limited on my available resources as far as tapes are concerned, and so I don't have the ability to maintain a stack of 60 unused tapes waiting to check out full ones and refill with empties. Two months will fill all my available tapes. I am basically needing to do a poor man's de-dupe, really, so that there is only one copy of a server's data in the library at any given time, even three, four or five years from now. Is that possible? Also, I am using spooling to maximize performance on the drives. I have the max bytes set to about 90% of my spool disks' capacity. However, I didn't see anywhere if there was a low bytes setting, so that if the pool fills, it will spool off to tape all the way down to the minimum instead of just until the spool says I'm not at my max capacity anymore. Or does it do that and I am just not setting the max size right and have it a decimal place too small? Any help is appreciated. And this Bacula thing is really a god-send. I am glad you said any. Although I try to help most users I am way too busy with the (paying) day job. So I can not spend more than a minute or 2 per reply... Anyways bacula prevents overwriting of volumes. You want to look at: http://www.bacula.org/en/rel-manual/Automatic_Volume_Recycling.html for ways to have bacula recycle your data based on the policies you want. John -- Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM) software. With Adobe AIR, Ajax developers can use existing skills and code to build responsive, highly engaging applications that combine the power of local resources and data with the reach of the web. Download the Adobe AIR SDK and Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users