In breaking changes 23.0.0 release note you will notice some information about the blocksize
https://docs.bareos.org/Appendix/ReleaseNotes.html#breaking-changes ;-) Maybe your drive or scsi adapter in not able to run 1MB On Sunday 14 July 2024 at 18:13:51 UTC+2 Alexandre Denault wrote: > Hi, > > So it took a quick bit of debugging, but I discovered that Debian had > changed my default block size to 512. Issuing > > mt -f /dev/nst0 defblksize 0 > > brought the drive back to variable block size, which fixed the problem. > > I can always add in more information if someone is curious. > > Cheers, > > On Sunday, July 14, 2024 at 12:01:18 AM UTC-4 Alexandre Denault wrote: > >> Hi, >> >> I've also noticed the following in dmesg: >> >> [st0] Incorrect block size. >> >> Could it be that Bareos is not reading the correct blocksize for the >> header of the tape? >> >> Cheers, >> >> On Saturday, July 13, 2024 at 9:21:50 PM UTC-4 Alexandre Denault wrote: >> >>> Hi, >>> >>> For the last few weeks now, I've been unable to mount any of my LTO-5 >>> drive. I always get the following error >>> >>> *stored/block.cc:1004 Read error on fd=6 at file:blk 0:0 on device >>> "TapeDrive0" (/dev/nst0). ERR=Input/output error.* >>> >>> At first, I suspected the drive was at fault. However, xTalk did not >>> detect any problem with the drive itself. >>> >>> >>> >>> >>> >>> >>> >>> >>> >>> >>> >>> >>> *Test Summary Drive Detection Test -------- : Passed SCSI >>> Interconnect Test ------ : Passed Tape Load Test -------------- : >>> Passed Quick Write/Read Test ------- : Passed Small Buffer >>> Wr/Rd Test ----- : Passed System Level Test ----------- : Passed >>> Time Test ------------------- : Passed Multi-Pattern Test >>> ---------- : Passed LTO Error Rate Test -------- : PassedNo Problem >>> DetectedThe tests did not detect any problems.* >>> >>> The test suite seems quite complete, including a read of 50Gb of data >>> and a validation read. >>> >>> I also tried with a different LTO5 drive, but got the same result. >>> >>> I formatted one my tape and re-labeled it. I was able to mount that tape >>> perfectly and complete some backups on it with no problems. However, all my >>> old backups on tapes remain inaccessible. Right now, it looks like I will >>> eventually have to reformat and relabel all my tapes to use them with >>> Bareos. I have over 50 tapes with 2 years of backups. >>> >>> This problem started around the time I updated to Bareos 23, although I >>> can't confirm that this is the problem. I am also unable to test >>> downgrading to Bareos 22 as I can't find any package for that. >>> >>> My drive is a a Quantum Ultrium 5 SAS (internal) and I use a mix of >>> Fujifilm and Quantum LTO 5 tapes. >>> >>> Director: 23.0.4~pre61.010c81fdc (03 July 2024) Debian GNU/Linux 11 >>> (bullseye) >>> Storage (tape): 23.0.4~pre74.8cb0a0c26 (10 July 2024) Debian GNU/Linux >>> 11 (bullseye) >>> Any suggestions to what I should tried to debug this problem? I would >>> rather not loose 2 years of backups if I can. >>> >>> Thanks, >>> >>> >>> >>> >> -- You received this message because you are subscribed to the Google Groups "bareos-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to bareos-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/bareos-users/42f7c2fb-b05b-4f07-8077-4e8d3b8fe2efn%40googlegroups.com.
