Re: [Bacula-users] Error - The number of files mismatch
You mean deleting volume from catalog? Check out delete command for console --http://www.bacula.org/rel-manual/Bacula_Console.html#SECTION000206000 JoeOn 4/10/06, Joe Park [EMAIL PROTECTED] wrote: Hello,http://www.bacula.org/rel-manual/Bacula_Freque_Asked_Questi.html#SECTION000348000 mt -f /dev/st0 rewindmt -f /dev/st0 weofIt will reset the tape. mt -f /dev/st0 erase will actually erase them (writing 0s I think) but it will take longer that just writing weof at the beginning of the tape. As for file number mismatch, have you read -- http://www.bacula.org/rel-manual/Tips_Suggestions.html#SECTION000357000 JoeOn 4/10/06, Bradley Schatz [EMAIL PROTECTED] wrote: Hi,I am a newcomer to using bacula, and am being stung by what must be pretty obvious to veterans. I have read the manual cover to cover in search of answers, but cant seem to figure this out. I am consistently having tapes marked with an error as follows: 11-Apr 08:23 foundry-sd: Volume Daily01 previously written, moving to end of data.11-Apr 08:23 foundry-sd: NightlySaveFoundry.2006-04-11_03.05.00 Error: I canot write on Volume Daily01 because: The number of files mismatch! Volume=27 Catalog=2811-Apr 08:23 foundry-sd: Marking Volume Daily01 in Error in Catalog.11-Apr 08:25 foundry-dir: Recycled volume Daily03What causes this kind of off by one error? And how can I completely reset the state of a particular volume to empty to start again? Thanks,bradley
Re: [Bacula-users] difference between autostarting and manually starting bacula daemon?
Hello,My machine is finally available (was doing stress testing) and tried modifing group parameters of bacula daemon start up scripts. I used operator group for both dir and sd daemon and confirmed it in ps. However, I still see same problem. After system boot up, tried to label a tape and got same error -- exited with code 127. I stopped the process and started the daemons manually, and this time, label command failed here too -- exited with code 1. Checking the device permission, I found that /dev/ch0 only had root user access (600). I added following to /etc/devfs.conf file -- own ch0 root:operatorperm ch0 0660Restarted the machine, confirmed /dev/ch0 has 660 with root:operator, confirmed bacula daemon parameters in ps. Then tried to label a tape in console and still same problem -- exited with code 127. Stopped the daemons, started them manually, and still same problem (exited with code 1). As a test, I tried following -- removed all setting in /etc/devfs.conf file and used -u root -g wheel parameters for sd and dir daemon. I restarted the machine and tried label again. And it threw same error -- exited with code 127. output from ps --root 723 0.0 0.7 6084 3396 ?? Ss 11:11AM 0:00.05 /usr/local/sbin/bacula-dir -u root -g wheel -v -c /usr/local/etc/bacula-dir.confroot 730 0.0 0.4 3864 2192 ?? Ss 11:11AM 0: 00.02 /usr/local/sbin/bacula-fd -u root -g wheel -v -c /usr/local/etc/bacula-fd.confroot 738 0.0 0.5 4804 2580 ?? Ss 11:11AM 0:00.06 /usr/local/sbin/bacula-sd -u root -g wheel -v -c /usr/local/etc/bacula- sd.conflabel failing --Automatically selected Pool: DefaultConnecting to Storage daemon DellPowervault_122T at 192.168.3.114:9103 ...Sending label command for Volume TestVol009 Slot 8 ... 3301 Issuing autochanger loaded drive 0 command.3991 Bad autochanger loaded drive 0 command: ERR=Child exited with code 127.3301 Issuing autochanger loaded drive 0 command. 3991 Bad autochanger loaded drive 0 command: ERR=Child exited with code 127.3304 Issuing autochanger load slot 8, drive 0 command.3992 Bad autochanger load slot 8, drive 0: ERR=Child exited with code 127. Label command failed for Volume TestVol009.Do not forget to mount the drive!!!I stopped the daemon and started them manually and label worked without any error. It just doesn't make sense... Any idea on this? Thanks for the help!JoeOn 4/6/06, Dan Langille [EMAIL PROTECTED] wrote: On 6 Apr 2006 at 18:22, Joe Park wrote: Ah I sse that.Thank you, I'll try that.If that works, or for that matter, does not, try similar for thedirector Joe On 4/6/06, Dan Langille [EMAIL PROTECTED] wrote: On 6 Apr 2006 at 18:10, Joe Park wrote:Hmm.. I don't have z-bacula.sh script -- only z-bacula-dir.sh , z-bacula-fd.sh, z-bacula-sd.sh scripts under /usr/local/etc/rc.d. I installed bacula 1.38.6 using latest ports snapshot.Maybe I should just download the source from bacula page and go from there? stick with the port. In /etc/rc.conf, add this: bacula_sd_flags= -u bacula -g operator -v -c /usr/local/etc/bacula- sd.conf and issue this command: /usr/local/etc/rc.d/z-bacula-sd.sh restart confirm you see -g operator here: ps auwx | grep bacula-sd Then try the autochanger.-- Dan Langille : Software Developer looking for work my resume: http://www.freebsddiary.org/dan_langille.php --Dan Langille : Software Developer looking for work my resume: http://www.freebsddiary.org/dan_langille.php
Re: [Bacula-users] column file.markedid does not exist error
Can you tell me any consequences from this bug? I'm testing bacula before using them in our production environment and would like to know if this is a show stopper.Thanks,Joe On 4/11/06, Kern Sibbald [EMAIL PROTECTED] wrote: On Monday 10 April 2006 21:33, Joe Park wrote: bacula ver: 1.38.6 OS: FreeBSD 6.0 DB: postgresql 8.1.3 Changer/Tapedrive: Dell PowerVault 122T Hello, I have this error messeges repeated in system log file -- Apr9 03:40:47 localhost postgres[18833]: [1-1] ERROR:column file.markedid does not exist Apr9 06:38:20 localhost postgres[20267]: [1-1] ERROR:column file.markedid does not exist Apr9 10:04:23 localhost postgres[21200]: [1-1] ERROR:column file.markedid does not exist The error coincides with bacula verify job end time.I see that bacula db has file.mark column, but not file.markedid column.Any help is appreciated.This is a bug (typo in SQL).I have just fixed it.Thanks for reporting it.The fix will be in 1.38.8--Best regards, Kern(/\V_V
[Bacula-users] column file.markedid does not exist error
bacula ver: 1.38.6OS: FreeBSD 6.0DB: postgresql 8.1.3Changer/Tapedrive: Dell PowerVault 122T Hello,I have this error messeges repeated in system log file --Apr 9 03:40:47 localhost postgres[18833]: [1-1] ERROR: column file.markedid does not existApr 9 06:38:20 localhost postgres[20267]: [1-1] ERROR: column file.markedid does not existApr 9 10:04:23 localhost postgres[21200]: [1-1] ERROR: column file.markedid does not existThe error coincides with bacula verify job end time. I see that bacula db has file.mark column, but not file.markedid column. Any help is appreciated.Thanks,Joe
[Bacula-users] column file.markedid does not exist error
bacula ver: 1.38.6OS: FreeBSD 6.0DB: postgresql 8.1.3Changer/Tapedrive: Dell PowerVault 122T Hello,I have this error messeges repeated in system log file --Apr 9 03:40:47 localhost postgres[18833]: [1-1] ERROR: column file.markedid does not existApr 9 06:38:20 localhost postgres[20267]: [1-1] ERROR: column file.markedid does not existApr 9 10:04:23 localhost postgres[21200]: [1-1] ERROR: column file.markedid does not existThe error coincides with bacula verify job end time. I see that bacula db has file.mark column, but not file.markedid column. Any help is appreciated.Thanks,Joe
Re: [Bacula-users] column file.markedid does not exist error
Nope, it was a fresh install of 1.38.6, using latest freebsd port snapshot.JoeOn 4/10/06, Ryan Novosielski [EMAIL PROTECTED] wrote:-BEGIN PGP SIGNED MESSAGE-Hash: SHA1 Did you upgrade to 1.38 from 1.36 without running the upgrade databasescript by any chance? __ __ __ |Y#| || |\/| |\ |\ || |Ryan Novosielski - User Support Spec. III |$| |__| || |__/ | \| _| |novosirj@ umdnj.edu - 973/972.0922 (2-0922) \__/ Univ. of Med. and Dent.|IST/AST - NJMS Medical Science Bldg - C630Joe Park wrote: bacula ver: 1.38.6 OS: FreeBSD 6.0 DB: postgresql 8.1.3 Changer/Tapedrive: Dell PowerVault 122T Hello, I have this error messeges repeated in system log file -- Apr9 03:40:47 localhost postgres[18833]: [1-1] ERROR:column file.markedid does not exist Apr9 06:38:20 localhost postgres[20267]: [1-1] ERROR:column file.markedid does not exist Apr9 10:04:23 localhost postgres[21200]: [1-1] ERROR:column file.markedid does not exist The error coincides with bacula verify job end time.I see that bacula db has file.mark column, but not file.markedid column.Any help is appreciated. Thanks, Joe-BEGIN PGP SIGNATURE-Version: GnuPG v1.4.2.2 (MingW32)iD8DBQFEOr49mb+gadEcsb4RAheDAKCxDbh4Dl+5/4UHOejPquIjKAJOHgCgvzehyOKMA4UNA5oOdvq2PkcXAvk= =M187-END PGP SIGNATURE-
Re: [Bacula-users] Error - The number of files mismatch
Hello,http://www.bacula.org/rel-manual/Bacula_Freque_Asked_Questi.html#SECTION000348000 mt -f /dev/st0 rewindmt -f /dev/st0 weofIt will reset the tape. mt -f /dev/st0 erase will actually erase them (writing 0s I think) but it will take longer that just writing weof at the beginning of the tape. As for file number mismatch, have you read --http://www.bacula.org/rel-manual/Tips_Suggestions.html#SECTION000357000 JoeOn 4/10/06, Bradley Schatz [EMAIL PROTECTED] wrote: Hi,I am a newcomer to using bacula, and am being stung by what must be pretty obvious to veterans. I have read the manual cover to cover in search of answers, but cant seem to figure this out. I am consistently having tapes marked with an error as follows: 11-Apr 08:23 foundry-sd: Volume Daily01 previously written, moving to end of data.11-Apr 08:23 foundry-sd: NightlySaveFoundry.2006-04-11_03.05.00 Error: I canot write on Volume Daily01 because: The number of files mismatch! Volume=27 Catalog=2811-Apr 08:23 foundry-sd: Marking Volume Daily01 in Error in Catalog.11-Apr 08:25 foundry-dir: Recycled volume Daily03What causes this kind of off by one error? And how can I completely reset the state of a particular volume to empty to start again? Thanks,bradley
[Bacula-users] difference between autostarting and manually starting bacula daemon?
Hello,I've been testing bacula for awhile and could not figure out this problem. Following is my environment --bacula ver: 1.38.6OS: FreeBSD 6.0DB: postgresql 8.1.3Changer/Tapedrive: Dell PowerVault 122T When I try something like label after boot up, autochanger throws error, as if it does not have correct permission for autochanger device --Connecting to Director localhost:9101 1000 OK: localhost-dir Version: 1.38.6 (28 March 2006) Enter a period to cancel a command.*labelUsing default Catalog name=MyCatalog DB=bacula The defined Storage resources are: 1: File 2: DellPowervault_122TSelect Storage resource (1-2): 2 Connecting to Storage daemon DellPowervault_122T at 192.168.3.114:9103 ...Enter new Volume name: TestVol04 Enter slot (0 or Enter for none): 1 Automatically selected Pool: Default Connecting to Storage daemon DellPowervault_122T at 192.168.X.XXX:9103 ...Sending label command for Volume TestVol04 Slot 1 ...3301 Issuing autochanger loaded drive 0 command.3991 Bad autochanger loaded drive 0 command: ERR=Child exited with code 127. 3301 Issuing autochanger loaded drive 0 command.3991 Bad autochanger loaded drive 0 command: ERR=Child exited with code 127.3304 Issuing autochanger load slot 1, drive 0 command. 3992 Bad autochanger load slot 1, drive 0: ERR=Child exited with code 127.Label command failed for Volume TestVol04.Do not forget to mount the drive!!!*qHowever, if I kill all three 3 daemons (either kill pid, or z-bacula-dir.sh stop) and start the daemons manually (/usr/local/etc/rc.d/z-bacula-dir.sh start, etc), the problem goes away. I made sure correct permission / ownership ( 0660 root:bacula ) is set on the devices and I don't see any differences (same options) when I look at the process with ps -- same whether or not daemon was started during boot time or manually by me. Any ideas? Thanks,Joe
[Bacula-users] difference between autostarting and manually starting bacula daemon?
Hello,I've been testing bacula for awhile and could not figure out this problem. Following is my environment --bacula ver: 1.38.6OS: FreeBSD 6.0DB: postgresql 8.1.3Changer/Tapedrive: Dell PowerVault 122T When I try something like label after boot up, autochanger throws error, as if it does not have correct permission for autochanger device --Connecting to Director localhost:9101 1000 OK: localhost-dir Version: 1.38.6 (28 March 2006) Enter a period to cancel a command.*labelUsing default Catalog name=MyCatalog DB=bacula The defined Storage resources are: 1: File 2: DellPowervault_122TSelect Storage resource (1-2): 2 Connecting to Storage daemon DellPowervault_122T at 192.168.3.114:9103 ...Enter new Volume name: TestVol04 Enter slot (0 or Enter for none): 1 Automatically selected Pool: Default Connecting to Storage daemon DellPowervault_122T at 192.168.X.XXX:9103 ...Sending label command for Volume TestVol04 Slot 1 ...3301 Issuing autochanger loaded drive 0 command.3991 Bad autochanger loaded drive 0 command: ERR=Child exited with code 127. 3301 Issuing autochanger loaded drive 0 command.3991 Bad autochanger loaded drive 0 command: ERR=Child exited with code 127.3304 Issuing autochanger load slot 1, drive 0 command. 3992 Bad autochanger load slot 1, drive 0: ERR=Child exited with code 127.Label command failed for Volume TestVol04.Do not forget to mount the drive!!!*qHowever, if I kill all three 3 daemons (either kill pid, or z-bacula-dir.sh stop) and start the daemons manually (/usr/local/etc/rc.d/z-bacula-dir.sh start, etc), the problem goes away. I made sure correct permission / ownership ( 0660 root:bacula ) is set on the devices and I don't see any differences (same options) when I look at the process with ps -- same whether or not daemon was started during boot time or manually by me. Any ideas? Thanks,
Re: [Bacula-users] difference between autostarting and manually starting bacula daemon?
Hello,Yup, autochanger passed the tests. Since it works when I manually start the daemon, it's not really permission issue no? I think I'll try changing group setting anyway.BTW, great article! I've been following your page closely for my bacula project and it's been a great help. Thanks! JoeOn 4/6/06, Dan Langille [EMAIL PROTECTED] wrote: On 6 Apr 2006 at 14:52, Joe Park wrote: Hello, I've been testing bacula for awhile and could not figure out this problem. Following is my environment -- bacula ver: 1.38.6 OS: FreeBSD 6.0 DB: postgresql 8.1.3Same here. Changer/Tapedrive: Dell PowerVault 122T When I try something like label after boot up, autochanger throws error, as if it does not have correct permission for autochanger device -- Connecting to Director localhost:9101 1000 OK: localhost-dir Version: 1.38.6(28 March 2006) Enter a period to cancel a command. *label Using default Catalog name=MyCatalog DB=bacula The defined Storage resources are:1: File2: DellPowervault_122T Select Storage resource (1-2): 2 Connecting to Storage daemon DellPowervault_122T at 192.168.3.114:9103 ... Enter new Volume name: TestVol04 Enter slot (0 or Enter for none): 1 Automatically selected Pool: Default Connecting to Storage daemon DellPowervault_122T at 192.168.X.XXX:9103 ... Sending label command for Volume TestVol04 Slot 1 ... 3301 Issuing autochanger loaded drive 0 command. 3991 Bad autochanger loaded drive 0 command: ERR=Child exited with code 127. 3301 Issuing autochanger loaded drive 0 command. 3991 Bad autochanger loaded drive 0 command: ERR=Child exited with code 127. 3304 Issuing autochanger load slot 1, drive 0 command. 3992 Bad autochanger load slot 1, drive 0: ERR=Child exited with code 127. Label command failed for Volume TestVol04. Do not forget to mount the drive!!! *q However, if I kill all three 3 daemons (either kill pid, or z-bacula-dir.sh stop) and start the daemons manually (/usr/local/etc/rc.d/z-bacula-dir.sh start, etc), the problem goes away. I made sure correct permission / ownership ( 0660 root:bacula ) is set on the devices and I don't see any differences (same options) when I look at the process with ps -- same whether or not daemon was started during boot time or manually by me.I generally keep the permissions on the devices unchanged.I thenalter the groups to suit.That is, my devices are root:operator.Atthe URL below, you can see the permissions. Did your autchanger pass the tests as described in the manual?Shortcuts can be found here where I documented my testing:http://www.freebsddiary.org/digital-tl891.php --Dan Langille : Software Developer looking for workmy resume: http://www.freebsddiary.org/dan_langille.php
