tape server on netra t1
Hi!
I am having problems with linux-sparc and tape
drives.
I set up a netra t1 as an amanda backup server,
which runs debian sid and kernel 2.6.5-rc3-bk4.
10 others linux boxes are the clients.
Minimal fs corruption occurs when the tape is
under heavy I/O.
The server has attached externally a dell PV-122T
tape loader and internally 2 scsi hard drives, the
loader and disks in different SCSI busses:
<--- BEGIN ---<
sym0: <896> rev 0x7 at pci 0000:02:08.0 irq 4,7e0
sym0: No NVRAM, ID 7, Fast-40, LVD, parity checking
sym0: SCSI BUS has been reset.
scsi0 : sym-2.1.18i
Vendor: SEAGATE Model: ST318305LSUN18G Rev: 0340
Type: Direct-Access ANSI SCSI revision: 03
sym0:0:0: tagged command queuing enabled, command queue depth 16.
sym0:0: FAST-40 WIDE SCSI 80.0 MB/s ST (25.0 ns, offset 31)
Vendor: FUJITSU Model: MAJ3182M SUN18G Rev: 0804
Type: Direct-Access ANSI SCSI revision: 02
sym0:1:0: tagged command queuing enabled, command queue depth 16.
sym1: <896> rev 0x7 at pci 0000:02:08.1 irq 4,7e0
sym1: No NVRAM, ID 7, Fast-40, LVD, parity checking
sym1: SCSI BUS has been reset.
scsi1 : sym-2.1.18i
Vendor: DELL Model: PV-122T Rev: D37r
Type: Medium Changer ANSI SCSI revision: 02
Vendor: HP Model: Ultrium 1-SCSI Rev: E32K
Type: Sequential-Access ANSI SCSI revision: 03
sym1:6: FAST-40 WIDE SCSI 80.0 MB/s ST (25.0 ns, offset 15)
st: Version 20040318, fixed bufsize 32768, s/g segs 256
Attached scsi tape st0 at scsi1, channel 0, id 6, lun 0
st0: try direct i/o: yes (alignment 512 B), max page reachable by HBA 98140
SCSI device sda: 35378533 512-byte hdwr sectors (18114 MB)
SCSI device sda: drive cache: write through
sda: sda1 sda2 sda3
Attached scsi disk sda at scsi0, channel 0, id 0, lun 0
sym0:1: FAST-40 WIDE SCSI 80.0 MB/s ST (25.0 ns, offset 31)
SCSI device sdb: 35378533 512-byte hdwr sectors (18114 MB)
SCSI device sdb: drive cache: write through
sdb: sdb1 sdb2 sdb3
Attached scsi disk sdb at scsi0, channel 0, id 1, lun 0
Attached scsi generic sg0 at scsi0, channel 0, id 0, lun 0, type 0
Attached scsi generic sg1 at scsi0, channel 0, id 1, lun 0, type 0
Attached scsi generic sg2 at scsi1, channel 0, id 5, lun 0, type 8
Attached scsi generic sg3 at scsi1, channel 0, id 6, lun 0, type 1
>--- END --->
Everything works as expected until the tape drive at /dev/nst0 is
under fire, when heavy tape I/O occurs I get the following errors,
which I have never seen before without the tape loader:
<--- BEGIN ---<
SABRE0: PCI SERR signal asserted.
SABRE0: PCI bus error, PCI_STATUS[2aa0]
SABRE0: Uncorrectable Error, primary error type[DMA Write:Translation Error]
SABRE0: bytemask[00ff] dword_offset[6] was_block(0)
SABRE0: UE AFAR [000000002f28b530]
SABRE0: UE Secondary errors [(DMA Write)(Translation Error)]
SABRE0: IOMMU Error, type[Invalid Error]
SABRE0: IOMMU TAG(8)[RAW(0000000000c616a6)error(Invalid Error)wr(0)sz(8K)vpg(c2d4c000)]
SABRE0: IOMMU DATA(8)[RAW(000000006fe00000)valid(1)used(1)cache(0)ppg(0000000000000000)
SABRE0: PCI SERR signal asserted.
SABRE0: PCI bus error, PCI_STATUS[0aa0]
SABRE0: PCI SERR signal asserted.
SABRE0: PCI bus error, PCI_STATUS[0aa0]
< 2 next lines repetitive... >
SABRE0: PCI SERR signal asserted.
SABRE0: PCI bus error, PCI_STATUS[0aa0]
>--- END --->
Then fs errors occurs and debian mount the fs as read-only, a reboot
and fsck.ext3 fixes the problem.
-otto
Reply to: