Re: [Bacula-users] RunAfter job problem
On Sunday 18 December 2005 05:46 am, Natxo Asenjo wrote: On Sat, 17 Dec 2005, 09:54:28PM +0100¨, Arno Lehmann said: 17-Dec 20:26 tux-dir: RunAfter: root 17-Dec 20:26 tux-dir: RunAfter: + /etc/bacula/delete_catalog_backup 17-Dec 20:26 tux-dir: RunAfter: + mt rewind 17-Dec 20:26 tux-dir: RunAfter: mt: /dev/tape: Permission denied 17-Dec 20:26 tux-dir: RunAfter: + mt eject 17-Dec 20:26 tux-dir: RunAfter: mt: /dev/tape: Permission denied 17-Dec 20:26 tux-dir: RunAfter: + exit 0 Hmm. Are you using a debian linux system? I think I remember something... yes, debian stable but bacula from sources. The problem is (well, for me in this case) that the jobs run as user bacula (this is good, do not get me wrong). The user bacula in debian systems (I used first the *.debs but then compiled) get a /bin/false shell. I changed that to /bin/bash. Then I edited the sudoers file and granted bacula access to /bin/mt as root without password. That did it. It is not beautiful but it works, and this machine is not a safety problem anyway. On my Debian Sarge system the tape device nodes are owned by 'root' with group 'tape'. Adding user 'bacula' to the 'tape' group in /etc/group solved the permissions problem. Of course, your changing /bin/false to /bin/bash is necessary, also! ;-) Cheers! cmr -- Debian 'Sarge': Registered Linux User #241964 More laws, less justice. -- Marcus Tullius Ciceroca, 42 BC --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_idv37alloc_id865op=click ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] RunAfter job problem
On Sat, 17 Dec 2005, 09:54:28PM +0100¨, Arno Lehmann said: 17-Dec 20:26 tux-dir: RunAfter: root 17-Dec 20:26 tux-dir: RunAfter: + /etc/bacula/delete_catalog_backup 17-Dec 20:26 tux-dir: RunAfter: + mt rewind 17-Dec 20:26 tux-dir: RunAfter: mt: /dev/tape: Permission denied 17-Dec 20:26 tux-dir: RunAfter: + mt eject 17-Dec 20:26 tux-dir: RunAfter: mt: /dev/tape: Permission denied 17-Dec 20:26 tux-dir: RunAfter: + exit 0 Hmm. Are you using a debian linux system? I think I remember something... yes, debian stable but bacula from sources. The problem is (well, for me in this case) that the jobs run as user bacula (this is good, do not get me wrong). The user bacula in debian systems (I used first the *.debs but then compiled) get a /bin/false shell. I changed that to /bin/bash. Then I edited the sudoers file and granted bacula access to /bin/mt as root without password. That did it. It is not beautiful but it works, and this machine is not a safety problem anyway. -- Groeten, J.I.Asenjo --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_idv37alloc_id865op=click ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
[Bacula-users] RunAfter job problem
hi, Following the instructions in http://bacula.org/rel-manual/Backup_Strategies.html#SECTION000283300 I have a little problem. The script /etc/bacula/end_of_backup.sh when run manually works fine, when run at the end of the job I get this message: 17-Dec 20:26 tux-dir: RunAfter: + echo root 17-Dec 20:26 tux-dir: RunAfter: root 17-Dec 20:26 tux-dir: RunAfter: + /etc/bacula/delete_catalog_backup 17-Dec 20:26 tux-dir: RunAfter: + mt rewind 17-Dec 20:26 tux-dir: RunAfter: mt: /dev/tape: Permission denied 17-Dec 20:26 tux-dir: RunAfter: + mt eject 17-Dec 20:26 tux-dir: RunAfter: mt: /dev/tape: Permission denied 17-Dec 20:26 tux-dir: RunAfter: + exit 0 So obviously the tape is not ejected and I need to eject it manually. ls -l /etc/bacula/end_of_backup.sh -rwsr-sr-- 1 root bacula 97 Dec 17 20:16 /etc/bacula/end_of_backup.sh I have even (OMG) tried setting the suid bit on the script to no avail. In fact, I have a echo statement at the beginnig of the script that confirms that it runs as root (echo $USER), so I am not sure what the problem is. This is bacula 1.38.2 and for the rest it's great! I sure would like to have this daily tape rotation because at our company we already have this rotation and one of the workers is used to this, so i'd rather not change it. -- Groeten, J.I.Asenjo --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] RunAfter job problem
Hello, On 12/17/2005 8:38 PM, Natxo Asenjo wrote: hi, Following the instructions in http://bacula.org/rel-manual/Backup_Strategies.html#SECTION000283300 I have a little problem. The script /etc/bacula/end_of_backup.sh when run manually works fine, when run at the end of the job I get this message: 17-Dec 20:26 tux-dir: RunAfter: + echo root 17-Dec 20:26 tux-dir: RunAfter: root 17-Dec 20:26 tux-dir: RunAfter: + /etc/bacula/delete_catalog_backup 17-Dec 20:26 tux-dir: RunAfter: + mt rewind 17-Dec 20:26 tux-dir: RunAfter: mt: /dev/tape: Permission denied 17-Dec 20:26 tux-dir: RunAfter: + mt eject 17-Dec 20:26 tux-dir: RunAfter: mt: /dev/tape: Permission denied 17-Dec 20:26 tux-dir: RunAfter: + exit 0 Hmm. Are you using a debian linux system? I think I remember something... So obviously the tape is not ejected and I need to eject it manually. Obviously. You might try inserting something like echo 'umount TAPE' | bconsole in your script before the mt eject. And perhaps mt rewoffl is also worth a try. ls -l /etc/bacula/end_of_backup.sh -rwsr-sr-- 1 root bacula 97 Dec 17 20:16 /etc/bacula/end_of_backup.sh I have even (OMG) :-) tried setting the suid bit on the script to no avail. Shouldn't work, as far as I know... In fact, I have a echo statement at the beginnig of the script that confirms that it runs as root (echo $USER), so I am not sure what the problem is. This is bacula 1.38.2 and for the rest it's great! I sure would like to have this daily tape rotation because at our company we already have this rotation and one of the workers is used to this, so i'd rather not change it. Good reason, I think. Arno -- IT-Service Lehmann[EMAIL PROTECTED] Arno Lehmann http://www.its-lehmann.de --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] RunAfter job problem
Hello Arno, Now we are very close to finding the problem. Can you send me the beginning of the log output that shows reserve.c:493 ... ? It should have 3 or four of those lines reserve.c:493 before it goes into the endless loop. That is the part that will tell me what combination of drives and switches is not finding a drive in your case. What is important is that I can see all the output beginning with the first reserve.c:493, continuing up to the point that it starts repeating over and over reserve.c:552 reserve.c:571, On Saturday 17 December 2005 21:54, Arno Lehmann wrote: Hello, On 12/17/2005 8:38 PM, Natxo Asenjo wrote: hi, Following the instructions in http://bacula.org/rel-manual/Backup_Strategies.html#SECTION0002833000 000 I have a little problem. The script /etc/bacula/end_of_backup.sh when run manually works fine, when run at the end of the job I get this message: 17-Dec 20:26 tux-dir: RunAfter: + echo root 17-Dec 20:26 tux-dir: RunAfter: root 17-Dec 20:26 tux-dir: RunAfter: + /etc/bacula/delete_catalog_backup 17-Dec 20:26 tux-dir: RunAfter: + mt rewind 17-Dec 20:26 tux-dir: RunAfter: mt: /dev/tape: Permission denied 17-Dec 20:26 tux-dir: RunAfter: + mt eject 17-Dec 20:26 tux-dir: RunAfter: mt: /dev/tape: Permission denied 17-Dec 20:26 tux-dir: RunAfter: + exit 0 Hmm. Are you using a debian linux system? I think I remember something... So obviously the tape is not ejected and I need to eject it manually. Obviously. You might try inserting something like echo 'umount TAPE' | bconsole in your script before the mt eject. And perhaps mt rewoffl is also worth a try. ls -l /etc/bacula/end_of_backup.sh -rwsr-sr-- 1 root bacula 97 Dec 17 20:16 /etc/bacula/end_of_backup.sh I have even (OMG) :-) : tried setting the suid bit on the script to no avail. Shouldn't work, as far as I know... In fact, I have a echo statement at the beginnig of the script that confirms that it runs as root (echo $USER), so I am not sure what the problem is. This is bacula 1.38.2 and for the rest it's great! I sure would like to have this daily tape rotation because at our company we already have this rotation and one of the workers is used to this, so i'd rather not change it. Good reason, I think. Arno -- Best regards, Kern ( /\ V_V --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
Re: [Bacula-users] RunAfter job problem
Hello, On 12/17/2005 11:13 PM, Kern Sibbald wrote: Hello Arno, Now we are very close to finding the problem. Damn good news. Something this day (or yesterday, rather) really needs :-| Can you send me the beginning of the log output that shows reserve.c:493 ... ? No. See below. It should have 3 or four of those lines reserve.c:493 before it goes into the endless loop. That is the part that will tell me what combination of drives and switches is not finding a drive in your case. What is important is that I can see all the output beginning with the first reserve.c:493, continuing up to the point that it starts repeating over and over reserve.c:552 reserve.c:571, Ok, following is the start of todays evening session with Bacula. I simply have no lines from reserve.c line 493... see for yourself. 495 is what comes closest, and I have marked the end of the initialization process with goblin:/etc/bacula # more /var/bacula/server/bacula-sd.log.16-Dec-22\:03 bacula-sd: stored_conf.c:613 Inserting director res: elf-monitor bacula-sd: stored_conf.c:613 Inserting device res: Wangtek 6525 bacula-sd: stored_conf.c:613 Inserting device res: HP DAT 0 bacula-sd: stored_conf.c:613 Inserting autochanger res: FastStor 4000 bacula-sd: stored_conf.c:613 Inserting device res: Quantum DLT 0 bacula-sd: stored_conf.c:613 Inserting director res: elf-monitor bacula-sd: stored_conf.c:613 Inserting device res: Wangtek 6525 bacula-sd: stored_conf.c:613 Inserting device res: HP DAT 0 bacula-sd: stored_conf.c:613 Inserting autochanger res: FastStor 4000 bacula-sd: stored_conf.c:613 Inserting device res: Quantum DLT 0 goblin-sd: jcr.c:113 read_last_jobs seek to 188 goblin-sd: jcr.c:120 Read num_items=10 goblin-sd: pythonlib.c:94 No script dir. prog=SDStartUp goblin-sd: stored.c:451 calling init_dev /var/bacula/ goblin-sd: dev.c:242 init_dev: tape=0 dev_name=/var/bacula/ goblin-sd: stored.c:453 SD init done /var/bacula/ goblin-sd: stored.c:451 calling init_dev /dev/nst0 goblin-sd: dev.c:242 init_dev: tape=2 dev_name=/dev/nst0 goblin-sd: stored.c:453 SD init done /dev/nst0 goblin-sd: stored.c:466 calling first_open_device Wangtek 6525 (/dev/nst0) goblin-sd: bnet_server.c:83 Addresses host[ipv4:0.0.0.0:9103] goblin-sd: dev.c:277 open dev: tape=2 dev_name=Wangtek 6525 (/dev/nst0) vol= mode=OPEN_READ_ONLY goblin-sd: dev.c:334 open dev: device is tape goblin-sd: dev.c:349 Try open Wangtek 6525 (/dev/nst0) mode=OPEN_READ_ONLY nonblocking=2048 goblin-sd: dev.c:380 openmode=3 OPEN_READ_ONLY goblin-sd: dev.c:393 open dev: tape 6 opened goblin-sd: label.c:71 Enter read_volume_label res=0 device=Wangtek 6525 (/dev/nst0) vol= dev_Vol=*NULL* goblin-sd: dev.c:645 rewind_dev res=0 fd=6 Wangtek 6525 (/dev/nst0) goblin-sd: label.c:138 Big if statement in read_volume_label goblin-sd: label.c:206 Compare Vol names: VolName= hdr=QIC-525-0046 goblin-sd: label.c:222 Copy vol_name=QIC-525-0046 Volume Label: Id: Bacula 1.0 immortal VerNo : 11 VolName : QIC-525-0046 PrevVolName : VolFile : 0 LabelType : VOL_LABEL LabelSize : 162 PoolName : QIC MediaType : DC6525 PoolType : Backup HostName : goblin Date label written: 25-Feb-2005 20:21 goblin-sd: label.c:227 Leave read_volume_label() VOL_OK goblin-sd: dev.c:645 rewind_dev res=0 fd=6 Wangtek 6525 (/dev/nst0) goblin-sd: stored.c:451 calling init_dev /dev/nst1 goblin-sd: dev.c:242 init_dev: tape=2 dev_name=/dev/nst1 goblin-sd: stored.c:453 SD init done /dev/nst1 goblin-sd: autochanger.c:247 Locking changer HP DAT goblin-sd: autochanger.c:218 run_prog: /etc/bacula/mtx-changer /dev/sg4 loaded 0 /dev/nst1 0 stat=0 result=0 goblin-sd: autochanger.c:256 Unlocking changer HP DAT goblin-sd: stored.c:466 calling first_open_device HP DAT 0 (/dev/nst1) goblin-sd: dev.c:277 open dev: tape=2 dev_name=HP DAT 0 (/dev/nst1) vol= mode=OPEN_READ_ONLY goblin-sd: dev.c:334 open dev: device is tape goblin-sd: autochanger.c:247 Locking changer HP DAT goblin-sd: autochanger.c:218 run_prog: /etc/bacula/mtx-changer /dev/sg4 loaded 0 /dev/nst1 0 stat=0 result=0 goblin-sd: autochanger.c:256 Unlocking changer HP DAT goblin-sd: dev.c:349 Try open HP DAT 0 (/dev/nst1) mode=OPEN_READ_ONLY nonblocking=2048 goblin-sd: dev.c:380 openmode=3 OPEN_READ_ONLY goblin-sd: dev.c:393 open dev: tape 7 opened goblin-sd: label.c:71 Enter read_volume_label res=0 device=HP DAT 0 (/dev/nst1) vol= dev_Vol=*NULL* goblin-sd: dev.c:645 rewind_dev res=0 fd=7 HP DAT 0 (/dev/nst1) goblin-sd: label.c:104 return VOL_NO_MEDIA: Couldn't rewind device HP DAT 0 (/dev/nst1): ERR=dev.c:679 Rewind e rror on HP DAT 0 (/dev/nst1). ERR=No medium found. goblin-sd: stored.c:451 calling init_dev /dev/nst2 goblin-sd: dev.c:242 init_dev: tape=2 dev_name=/dev/nst2 goblin-sd: stored.c:453 SD init done /dev/nst2 goblin-sd: autochanger.c:247 Locking changer FastStor 4000 goblin-sd: autochanger.c:218 run_prog:
