Robin Blanchard wrote: > I've been around the block with LSI and with adaptec: tried an LSI > SYMC101, an adaptec 2940U2W and a 39160. I've removed the > library/exchange from the equation, using only the LTO-3 drive (and have > actually swapped that drive out for another as well), swapped SCSI > cables, and terminators, and still am getting scsi errors. Anyone got > any tips/ideas here ?
To be honest, this looks like a termination problem (or, at least, a bus problem of some sort). However, since you appear to have swapped out every piece of hardware, the only thing left seems to be the configuration .... Both the external and the internal segments of the SCSI bus need to be properly terminated. If not, errors will occur on the bus and the HBA will step back the clock speed in an attempt to make the bus work reliably. So far so good - you are seeing parity errors and the HBA is stepping the speed back from 160 MB/s to 40 MB/s. It looks as if there is either insufficient termination, or too much termination, on the bus - too much termination can be as bad as no termination. There should be configuration options in the HBA's BIOS to set termination of the internal and external segments at the HBA. Usually termination of the external and internal segments are configured independently. They can usually be set to auto (which is usually the default and the manufacturer's recommended setting), or explicitly on or off. Check you do not have termination switched on at the HBA *and* a terminator at the end of the cable. Is the LTO3 drive internal or external, BTW? The internal segment of the bus needs to follow the same rules as the external segment, but is usually more difficult to get right - termination can occur at the HBA, the drive(s) or the end of the cable. Normally for an Ultra 160 SCSI bus, the internal segment should have termination switched off (or set to auto) at the HBA, the devices on the bus should be unterminated and the ribbon cable should be terminated. Similarly (but less likely), check there is not an option in the LTO3 drive to terminate the bus. If the drive is terminating the bus, *and* there is a terminator screwed onto the back, this will add up to too much termination. Like I said, this is unlikely. If anything, if the drive can terminate the bus, it will most probably be automatic. If you have a mixture of wide and narrow devices on the bus (particularly the internal segment) termination can get a bit tricky :-) I have never used any of the 3 HBAs you mention at the start of the email, but I have used a 2940N and a 29160N. A couple of points - the 2940N did not (as far as I can remember) have an "auto" setting for bus termination at the HBA; it could only be set to "on" or "off". The 2940UW may be the same. Also, the 29160 BIOS tests the bus segments and reports if it detects a termination problem. The 39160 is, I believe, similar to the 29000 series, being mainly a dual-channel version. It may tell you of a termination problem if you look carefully at the BIOS output (and try not to blink at the wrong time in case you miss it :-) Depending on how the HBA works, it is possible that a termination error on one segment will affect both segments. If, for example, you have an external device attached, but no internal devices, then the HBA should have termination *on* for the internal segment, and *off* for the external segment. [Or "auto", of course.] Allan > target2:0:1: FAST-40 WIDE SCSI 80.0 MB/s ST (25 ns, offset 31) > Vendor: IBM Model: ULTRIUM-TD3 Rev: 73P5 > Type: Sequential-Access ANSI SCSI revision: 03 > target2:0:1: Beginning Domain Validation > target2:0:1: asynchronous > target2:0:1: wide asynchronous > target2:0:1: FAST-80 WIDE SCSI 160.0 MB/s DT (12.5 ns, offset 62) > sym0: SCSI parity error detected: SCR1=1 DBC=15000000 SBCL=2d > sym0: SCSI parity error detected: SCR1=1 DBC=15000000 SBCL=2d > target2:0:1: FAST-80 WIDE SCSI 160.0 MB/s DT (12.5 ns, offset 62) > sym0: SCSI parity error detected: SCR1=1 DBC=15000000 SBCL=2d > sym0: SCSI parity error detected: SCR1=1 DBC=15000000 SBCL=2d > target2:0:1: Domain Validation detected failure, dropping back > target2:0:1: control msgout: c. > sym0: TARGET 1 has been reset. > target2:0:1: Domain Validation detected failure, dropping back > target2:0:1: FAST-40 WIDE SCSI 66.0 MB/s DT (30.3 ns, offset 62) > sym0: SCSI parity error detected: SCR1=1 DBC=15000000 SBCL=2d > sym0: SCSI parity error detected: SCR1=1 DBC=15000000 SBCL=2d > target2:0:1: FAST-40 WIDE SCSI 66.0 MB/s DT (30.3 ns, offset 62) > sym0: SCSI parity error detected: SCR1=1 DBC=15000002 SBCL=ae > sym0: SCSI parity error detected: SCR1=1 DBC=15000000 SBCL=2d > sym0:1: ERROR (0:8) (d-ae-5) (3e/38/8c) @ (scripta 98:870b0000). > sym0: script cmd = 800a0000 > sym0: regdump: da 10 c0 38 47 3e 01 0a 03 0d 00 ae 80 00 0e 00 00 1c 6f > 08 2a 00 00 00. > sym0: SCSI BUS reset detected. > sym0: SCSI BUS has been reset. > target2:0:1: Domain Validation detected failure, dropping back > target2:0:1: FAST-20 WIDE SCSI 40.0 MB/s DT (50 ns, offset 62) > target2:0:1: FAST-20 WIDE SCSI 40.0 MB/s DT (50 ns, offset 62) > target2:0:1: FAST-20 WIDE SCSI 40.0 MB/s DT (50 ns, offset 62) > target2:0:1: FAST-20 WIDE SCSI 40.0 MB/s DT (50 ns, offset 62) > target2:0:1: FAST-20 WIDE SCSI 40.0 MB/s DT (50 ns, offset 62) > target2:0:1: Domain Validation skipping write tests > target2:0:1: Ending Domain Validation > scsi 2:0:1:0: Attached scsi generic sg2 type 1 > ACPI: PCI Interrupt 0000:05:02.1[B] -> GSI 101 (level, low) -> IRQ 22 > sym1: <1010-33> rev 0x1 at pci 0000:05:02.1 irq 22 > sym1: Symbios NVRAM, ID 7, Fast-80, SE, parity checking > sym1: open drain IRQ line driver, using on-chip SRAM > sym1: using LOAD/STORE-based firmware. > sym1: handling phase mismatch from SCRIPTS. > sym1: SCSI BUS has been reset. > scsi3 : sym-2.2.3 > ---------------------------------------- > Robin P. Blanchard > Systems Administrator > Information Technology Outreach Services > Carl Vinson Institute of Government > The University of Georgia > fon 706.542.6295 // fax 706.542.6535 > ---------------------------------------- > > ------------------------------------------------------------------------- > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for > just about anything Open Source. > http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace > _______________________________________________ > Bacula-users mailing list > Bacula-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/bacula-users ------------------------------------------------------------------------- Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace _______________________________________________ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
