Dear Julien, Under 2.6 try (note stsetoptions, not stoptions):
mt -f /dev/nst0 stsetoptions 0 dmesg | grep st0 That should give you the information you need. Clearly from the existing info your drive is coming Up incorrectly. What remains to be seen is whether we can force it permanently to block size of zero. Note that for 2.6 kernel, depending upon version of mt you should be able to: mt -f /dev/nst0 defblksize 0 as well. Regards, Brett -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Julien Cigar Sent: 14 July 2006 16:08 To: bacula-users@lists.sourceforge.net Subject: Re: [Bacula-users] Last chance before I [EMAIL PROTECTED] tapedrivethrough the window ... Brett, (Under Linux it's "stoptions" in place of "setstoptions".) Under kernel 2.6 I have no results in /var/log/syslog or /var/log/messages Under kernel 2.4 I have this: Jul 14 11:30:32 localhost kernel: st0: Mode 0 options: buffer writes: 1, async writes: 1, read ahead: 1 Jul 14 11:30:32 localhost kernel: st0: can bsr: 0, two FMs: 0, fast mteom: 0, auto lock: 0, Jul 14 11:30:32 localhost kernel: st0: defs for wr: 0, no block limits: 0, partitions: 0, s2 log: 0 Jul 14 11:30:32 localhost kernel: st0: sysv: 0 nowait: 0 When I boot (under kernel 2.6) I noticed: Jul 14 15:36:49 localhost kernel: scsi0 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 7.0 Jul 14 15:36:49 localhost kernel: <Adaptec 2940 Ultra SCSI adapter> Jul 14 15:36:49 localhost kernel: aic7880: Ultra Wide Channel A, SCSI Id=7, 16/253 SCBs Jul 14 15:36:49 localhost kernel: Jul 14 15:36:49 localhost kernel: Vendor: SONY Model: SDX-500C Rev: 0204 Jul 14 15:36:49 localhost kernel: Type: Sequential-Access ANSI SCSI revision: 02 Jul 14 15:36:49 localhost kernel: target0:0:6: Beginning Domain Validation Jul 14 15:36:49 localhost kernel: target0:0:6: wide asynchronous. Jul 14 15:36:49 localhost kernel: target0:0:6: FAST-10 WIDE SCSI 20.0 MB/s ST (100 ns, offset 8) Jul 14 15:36:49 localhost kernel: target0:0:6: Domain Validation skipping write tests Jul 14 15:36:49 localhost kernel: target0:0:6: Ending Domain Validation Jul 14 15:36:49 localhost kernel: st: Version 20050830, fixed bufsize 32768, s/g segs 256 Jul 14 15:36:49 localhost kernel: st 0:0:6:0: Attached scsi tape st0<4>st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 1048575 As you can see the drive starts with a fixed block size of 32k. So after the boot, I stop all the bacula daemons and I setblk 0 and stoptions 7 (buffer write + async writes + read ahead). Then I relaunch the bacula daemons. Is it possible that the kernel or the driver (or ...) "resets" the block size after a mt -f /dev/nst0 setblk 0 and an ejecting of the tape from the tape drive ? (I checked with mt -f /dev/nst0 status but it's still on '0' so I presume that the drive is still in variable block mode). Another question, is it possible that it fails because I don't *umount before ejecting the tape ? (I press on the eject button of the tape drive) In fact the people who have to change the tape when they received a mail should not access bconsole (too complicated ...), so all they have to do is to press on the button and put the other tape (so without having to *umount). Many thanks, Julien [EMAIL PROTECTED] wrote: > Dear Julian, > > This is good, but we were both referring to in between jobs (after daemon has > started). The original commands I gave: > mt -f /dev/nst0 setstoptions 0 > grep st0 /var/log/messages > > Will show you if these settings are being changed indvertedly by > either (a) bacula or (b) the kernel driver prior to the restore command (e.g. > by the mount command which 'opens' the tape drive). > > Best Regards, > > Brett > -----Original Message----- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of > Julien Cigar > Sent: 14 July 2006 15:38 > To: bacula-users@lists.sourceforge.net > Subject: Re: [Bacula-users] Last chance before I throw [EMAIL PROTECTED] > tapedrivethrough the window ... > > In fact I always do this _before_ launching the bacula storage daemon : > mt -f /dev/nst0 stoptions 7 > mt -f /dev/nst0 setblk 0 > > > Georger Araujo wrote: > >> Judging from the "BB02" string, I suspect it to be a problem with the >> block size. I've had such problems when I first deployed Bacula, so I >> ALWAYS set the block size before I mount a tape, and also after I >> eject it. >> Here's my set_blocksize.sh script: >> >> 8<--cut-here---------------- >> #!/bin/sh >> # Set variable block mode (BlockSize = 0) # Fixed block (BlockSize <> >> 0) is WRONG! >> # If in doubt, run tapeinfo -f /dev/sg0 - BlockSize should be 0 mt -f >> /dev/nst0 defblksize 0 # Rewind the tape mt -f /dev/nst0 rewind >> ---------------cut-here-->8 >> >> Here's my eject_tape.sh script: >> >> 8<--cut-here---------------- >> #! /bin/sh >> cd /etc/bacula >> # Set variable block mode >> # If in doubt, run tapeinfo -f /dev/sg0 - should be 0 mt -f /dev/nst0 >> defblksize 0 # Rewind the tape mt -f /dev/nst0 rewind # Eject the >> tape mt -f /dev/nst0 eject exit 0 >> ---------------cut-here-->8 >> >> While I'm at it, here's tape_properties.sh ... >> >> 8<--cut-here---------------- >> #! /bin/sh >> tapeinfo -f /dev/sg0 >> exit 0 >> ---------------cut-here-->8 >> >> ... and clear_label.sh: >> >> 8<--cut-here---------------- >> #!/bin/sh >> # Manually relabel a tape. >> # >> # Run bconsole >> # unmount >> # delete media volume >> >> # Set variable block mode >> # If in doubt, run tapeinfo -f /dev/sg0 - should be 0 mt -f /dev/nst0 >> defblksize 0 # Rewind the tape mt -f /dev/nst0 rewind # Write >> physical End Of File (EOF) to the tape mt -f /dev/nst0 weof >> >> # Run bconsole >> # label >> # mount >> ---------------cut-here-->8 >> >> My Bacula server (director/client/storage): Dell 2800 running RHELAS >> 4 with an IBM LTO-2 tape drive. >> Regards, >> >> Georger >> >> --- [EMAIL PROTECTED] escreveu: >> >> >> >>> Dear Julien, >>> >>> The only thing I can suggest now is that you review what happens to >>> the tape options via: >>> >>> mt -f /dev/nst0 stoptions 0 >>> grep st0 /var/log/messages >>> >>> Between backup and restore (i.e. before backup and before restore >>> but after the mount command) in case they are being reset somehow >>> (by the mount maybe?). >>> >>> The manual indicates that the error you are getting is possibly >>> related to block positioning. As you have indicated it might also >>> be a kernel issue. >>> >>> Best Regards, >>> >>> Brett >>> >>> >> >> >> >> >> >> >> _______________________________________________________ >> Você quer respostas para suas perguntas? Ou você sabe muito e quer >> compartilhar seu conhecimento? Experimente o Yahoo! Respostas ! >> http://br.answers.yahoo.com/ >> >> >> --------------------------------------------------------------------- >> - >> --- Using Tomcat but need to do more? Need to support web services, >> security? >> Get stuff done quickly with pre-integrated technology to make your >> job easier Download IBM WebSphere Application Server v.1.0.1 based on >> Apache Geronimo >> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121 >> 6 >> 42 _______________________________________________ >> Bacula-users mailing list >> Bacula-users@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/bacula-users >> >> > > > > ---------------------------------------------------------------------- > --- Using Tomcat but need to do more? Need to support web services, > security? > Get stuff done quickly with pre-integrated technology to make your job > easier Download IBM WebSphere Application Server v.1.0.1 based on > Apache Geronimo > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=1216 > 42 _______________________________________________ > Bacula-users mailing list > Bacula-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/bacula-users > > ______________________________________________________________________ > This email has been scanned by the MessageLabs Email Security System. > For more information please visit http://www.messagelabs.com/email > ______________________________________________________________________ > > ______________________________________________________________________ > This email has been scanned by the MessageLabs Email Security System. > For more information please visit http://www.messagelabs.com/email > ______________________________________________________________________ > ------------------------------------------------------------------------- Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 _______________________________________________ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users ______________________________________________________________________ This email has been scanned by the MessageLabs Email Security System. For more information please visit http://www.messagelabs.com/email ______________________________________________________________________ ______________________________________________________________________ This email has been scanned by the MessageLabs Email Security System. For more information please visit http://www.messagelabs.com/email ______________________________________________________________________ ------------------------------------------------------------------------- Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 _______________________________________________ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
