Operating System - OpenVMS
1748219 Members
4834 Online
108759 Solutions
New Discussion юеВ

Booting Standalone Backup tape from TZ88

 
Dan Mellem
Advisor

Re: Booting Standalone Backup tape from TZ88

Steven Schweda:

> Still wondering.

It only says TZ88 on the front on the drive itself and I don't want to pull the drive apart to look for some other part number. It's in the case from a different model and the bottom label says "MODEL 15/30 GB DLT" and "Series 3305" but that's likely not correct.

> Your battery's dead, too, I see.

When the battery is dead it also loses the BFLG and BOOT variables and asks for the time, unless there are two batteries. We replaced the battery about a year ago.

>  Apparently, someone hates something about your tape drive.  Time to buy more (experimental) junk?

I'm not sure what you're calling experimental. This drive is from DEC and is supported on the VAX.

Steven Schweda
Honored Contributor

Re: Booting Standalone Backup tape from TZ88

> [...] We replaced the battery about a year ago.

   If the battery's ok, then ">>>SET LANGUAGE" might save you a few
seconds per year.  (In my case, the system is almost always powered off,
so even a new battery will typically discharge between uses.  I also
have an IBM workstation which kills its oversized Li coin cell in less
time than it normally stays powered off.)

>    Apparently, someone hates something about your tape drive.  Time to
> buy more (experimental) junk?
>
> I'm not sure what you're calling experimental. This drive is from DEC
> and is supported on the VAX.

   Fine, but I assume that if you actually had support, then you'd be
running VMS V7.3 instead of V6.2, and you'd be complaining to your
support provider instead of asking here.

   If your drive really is supported, and it works for many other
purposes (but not this one), then I'd say that any hardware replacement
will be an experiment.  In this case, anything will be an experiment.
Not knowing why you get the failure makes it tough to be confident that
any other drive (supported or not) would do any better.  Only one way to
find out.  A quick look at Ebay suggests that there's no shortage of
claimed-working DEC/Compaq/Quantum/... 4000/7000/8000 drives in the
sub-$50 neighborhood, so trying another drive would not cost much.

   Personally, I'd probably skip ahead to a 7000 or 8000, but, as my
experience with V6.2 STABACKIT.COM and a 7000 drive suggests, it might
be safer to stick with drives which existed when your VMS was minted.

   For the record, the drive which I tested says that it's a Compaq
TH6AE-HK (in a Compaq box).  The box has 68-pin (wide) connectors.  It
had a wide (SE) terminator and a 68-to-50(small) cable, which led to a
CD-RW drive (powered off), thence to a 50(small)-to-50(large) cable, to
a home-made 50(large)-to-50(large) coupler, to a 50(large)-to-50(large)
cable to the KZQSA port.  (If that mess works, then what wouldn't?  That
SCSI stuff is practically fool-proof, right?)

Volker Halle
Honored Contributor

Re: Booting Standalone Backup tape from TZ88

Dan,

chapter 4.7.2 of the following manual (KA52 CPU System Maintenance) has all the details:

http://decdoc.itsx.net/dec94mds/473aamga.pdf

And your tape boot fails exactly here:

9. Copy the secondary bootstrap from the boot device into local memory above the stack. If this fails, the bootstrap fails.

The code executing here is VMB from the VAX 4000 firmware.

Does the tape move at all, when trying to boot ?

Volker.

Dan Mellem
Advisor

Re: Booting Standalone Backup tape from TZ88

Steven Schweda:

> Fine, but I assume that if you actually had support, then you'd be running VMS V7.3 instead of V6.2, and you'd be complaining to your support provider instead of asking here.

We actually do have support for both hardware and software, but we're running an ancient business system that only listed 6.2 as the highest supported OS. I ask here because I usually get more experienced people here than at the service center. The guys that really knew their stuff there have retired.

> That SCSI stuff is practically fool-proof, right?

:) So many SCSI headaches over the years. But this has been pretty stable, knock on wood.


 Volker Halle:

Thank you for the background and manual.

> Does the tape move at all, when trying to boot ?

The tape activity light flashes for a few seconds when I try to boot, but not a particularly long time.

Steven Schweda
Honored Contributor

Re: Booting Standalone Backup tape from TZ88

> > Does the tape move at all, when trying to boot ?
>
> The tape activity light flashes for a few seconds when I try to boot,
> but not a particularly long time.

   For comparison, around here, typical times look like these (mm:ss):

>>>boot mka400             T = 00:00
(BOOT/R5:0 MKA400)         T = 00:00

  2..                      T = 00:04  Tape activity starts.
-MKA400                    T = 00:10
  1..0..                   T = 00:14

   The last two times vary by a few seconds, (I assume) because the tape
wasn't always rewound between trials.  (The 4s difference is pretty
reliable.)

Dan Mellem
Advisor

Re: Booting Standalone Backup tape from TZ88

Those times are consistent with what I'm seeing, except instead of the "1..0.." countdown I get the "?56 DRVERR, MKA500" message.

Bob Blunt
Respected Contributor

Re: Booting Standalone Backup tape from TZ88

Dan, um...  I've been checking my own memory and what information I have at hand.  The KZQSA is a REALLY slow interface.  REALLY, really slow.  It also has some idiosyncracies that you may be running into.  From what I can tell the TZ88 USUALLY works but I can't find that it was actually officially supported on that controller.  There were also some driver concerns about the late V5.5 timeframe that may have not been fully resolved by the O/S release you're running.  Whatever "*bleep*" is?  We'd need to see ALL of the errorlog and not just tape(s).  Issues could/would be reported for either (or both) MKA500 and PKA0 IF there are any.  The VAX4000-600 with the KZQSA can also be a unique config, there's a SCSI I/F on the backplane that you have to connect to the KZQSA to use any of the internal SCSI bays (and, I presume, the SCSI connector that I'm pretty sure lives in the upper left corner of the front of the chassis).  So cabling can be...interesting.  It could also be cabled directly (and I'm trying to visualize the KZQSA-SA module for the BA4xx chassis and remember if it has two old style SCSI connectors on the bulkhead) from the SCSI connector on the bulkhead of the module to the tape drive(s).  If your tapes are connected that way, directly, cumulative cable length might be an issue.  I'd also try monkeying around with the device order and numbering on that bus to see if that helps any (put the TZ88 first, number it as MKA0, terminate it at the back of the TZ88, etc.

You also mentioned that the front of the drive says TZ88 and something underneath hinted at a different capacity (15/30?).  The tape drive desktop-ish boxes aren't that magical unless there's some logic internal to the box.  I've built a wide number of what vendors would call "customized" solutions and as long as the power being sent to the drive is correct and there aren't any funky SCSI adapters or logic being used to connect to the SCSI connectors on the back they USUALLY work just fine.  The ID on the front of the drive itself that the "world" sees will be the tape drive model but you'll find that there are a huge number of "bricks" that can be used inside.  The TZ88 model has been around quite a while (I found references older than 1996 to the drive) and they do tend to get rather picky as they age.  Those drives typically have three main drive logic cards, two that handle the drive and whatnot and the SCSI interface.  The SCSI interface module can, probably, be several different flavors like Single Ended, Fast, Differential, etc.  So I'm kinda curious what type connector(s) are on the back of the box containing this TZ88?  Are they the really old style SCSI 50 pin centronics type with bail locks (about 2" or so) or a smaller, denser connector?

Connector to the right above the ID selectorConnector to the right above the ID selector

 

There's also something banging around in my head about some odd behavior I experienced with building the S/A backup kit on certain tape drives.  I had to finagle around with the code, for instance, to get a successful build and boot from DAT drives...but that's been a LOOOOONG time ago (but I did get it working...and I don't recall how it failed before it did).  I won't be able to dig into that until after work, possibly as late as Monday.

bob

Steven Schweda
Honored Contributor

Re: Booting Standalone Backup tape from TZ88

> [...] The KZQSA is a REALLY slow interface.  REALLY, really slow.
> [...]

   Which, as I recall, is the main reason it was supported for CD-ROM
and tape, but not for hard disk.  (Although I have attached a hard disk,
long ago, and experienced no actual trouble.)

> Whatever "*bleep*" is?

   Short for "analyze".  This forum software is excessively prudish.
(Among other things.)

> there's a SCSI I/F on the backplane that you have to connect to the
> KZQSA to use any of the internal SCSI bays [...]

   From:

> [...] It's in the case from a different model [...]

I inferred an external box.  (Like mine.)

> [...] So cabling can be...interesting. [...]

   Or very boring.  Straight from the external box to a connector on the
bulkhead of the KZQSA, for example.  It's tough to believe that anyone
has a longer/messier arrangement than mine, however.

> [...] and remember if it has two old style SCSI connectors on the
> bulkhead) [...]

   Yup.  If the card is at one end of the bus, then (I believe that) the
second connector should get a terminator for that end of the bus.  (Mine
has a genuine Digital H8574-A ("REV A") in the lower hole, and the
probably-too-long, multi-piece cable in the upper hole.)

   Part of the mystery here is that the stuff is claimed to work for
everything but bootstrap.

> There's also something banging around in my head [...]

   As I said, I needed to add code for "DEV_TYPE .EQ. 28" to get the
V6.2 STABACKIT.COM to believe that my DLT7000 is a tape drive.  I doubt
that I was the first one to do that.  But a TZ88 was already supported
there, and the original complaint wasn't about making the kit, only
booting from it.

> [...] (but I did get it working...and I don't recall how it failed
> before it did).

   It's scrolled away here, but it involved some strange behavior,
followed by some odd-ball complaint about MOUNT qualifiers (because the
tape drive was believed to be a disk drive, I assume).  If anyone cares,
I could fire up the VAX, and publish the DIFFERENCES, but it's unlikely
to matter in this case until someone else installs a tape drive newer
than a TZ88.

   I don't know what you can do about the banging.

Chris Scheers
Advisor

Re: Booting Standalone Backup tape from TZ88

Perhaps related, perhaps not. When I was supporting a VAXstation 4000-90 with DAT running VMS 5.5-2, I kept a 7.3 STABACKIT DAT tape because the V5.5-2 STABACKIT did not boot correctly. I don't remember what error I got, but I know V5.5-2 did not work and V7.3 did.
Dan Mellem
Advisor

Re: Booting Standalone Backup tape from TZ88

Bob Blunt:

> Whatever "*bleep*" is?

Sorry, it's analyze, from analyze/error, but I normally type the first four letters of commands, which must get replaced in this forum.

> We'd need to see ALL of the errorlog and not just tape(s).  Issues could/would be reported for either (or both) MKA500 and PKA0 IF there are any.

I did analyze/error/since=2-mar-2017 (last time I rebooted the system) and they were just the normal startup messages:

 ******************************* ENTRY    9246. *******************************
 ERROR SEQUENCE 83.                              LOGGED ON:        SID 13000202
 DATE/TIME  2-MAR-2017 09:26:54.29                            SYS_TYPE 01460701
 SYSTEM UPTIME: 0 DAYS 00:01:16
 SCS NODE: VAX1                                                VAX/VMS V6.2

 SYSTEM START-UP  KA690-AA  CPU Microcode Rev # 2.  CONSOLE FW REV# 4.6
                      Standard Microcode Patch    Patch Rev # 1.

 TIME OF DAY CLOCK     2F1A0ED4
 ******************************* ENTRY    9247. *******************************
 ERROR SEQUENCE 84.                              LOGGED ON:        SID 13000202
 DATE/TIME  2-MAR-2017 09:26:54.77                            SYS_TYPE 01460701
 SYSTEM UPTIME: 0 DAYS 00:01:16
 SCS NODE: VAX1                                                VAX/VMS V6.2

 MOUNT VOLUME  KA690-AA  CPU Microcode Rev # 2.  CONSOLE FW REV# 4.6
                      Standard Microcode Patch    Patch Rev # 1.

       UNIT _SYSTEM$DIA0:, VOLUME LABEL "VAXVMSV055"

       419. QIO OPERATIONS THIS UNIT, 0. ERRORS THIS UNIT
 ******************************* ENTRY    9248. *******************************
 ERROR SEQUENCE 85.                              LOGGED ON:        SID 13000202
 DATE/TIME  2-MAR-2017 09:27:26.09                            SYS_TYPE 01460701
 SYSTEM UPTIME: 0 DAYS 00:01:48
 SCS NODE: VAX1                                                VAX/VMS V6.2

 MOUNT VOLUME  KA690-AA  CPU Microcode Rev # 2.  CONSOLE FW REV# 4.6
                      Standard Microcode Patch    Patch Rev # 1.

       UNIT _DISK2$DIA2:, VOLUME LABEL "DATA1"

       13. QIO OPERATIONS THIS UNIT, 0. ERRORS THIS UNIT
...

And the nightly backups don't show any errors, either:

******************************* ENTRY    9294. *******************************
 ERROR SEQUENCE 200.                             LOGGED ON:        SID 13000202
 DATE/TIME  2-MAR-2017 23:00:19.90                            SYS_TYPE 01460701
 SYSTEM UPTIME: 0 DAYS 11:45:45
 SCS NODE: VAX1                                                VAX/VMS V6.2

 MOUNT VOLUME  KA690-AA  CPU Microcode Rev # 2.  CONSOLE FW REV# 4.6
                      Standard Microcode Patch    Patch Rev # 1.

       UNIT _VAX1$MKA500:, VOLUME LABEL "BACKUP"

       35. QIO OPERATIONS THIS UNIT, 0. ERRORS THIS UNIT
 ******************************* ENTRY    9295. *******************************
 ERROR SEQUENCE 225.                             LOGGED ON:        SID 13000202
 DATE/TIME  3-MAR-2017 03:06:08.53                            SYS_TYPE 01460701
 SYSTEM UPTIME: 0 DAYS 15:51:34
 SCS NODE: VAX1                                                VAX/VMS V6.2

 TIME STAMP  KA690-AA  CPU Microcode Rev # 2.  CONSOLE FW REV# 4.6
                      Standard Microcode Patch    Patch Rev # 1.
 ******************************* ENTRY    9296. *******************************
 ERROR SEQUENCE 226.                             LOGGED ON:        SID 13000202
 DATE/TIME  3-MAR-2017 03:09:34.98                            SYS_TYPE 01460701
 SYSTEM UPTIME: 0 DAYS 15:55:00
 SCS NODE: VAX1                                                VAX/VMS V6.2

 DISMOUNT VOLUME  KA690-AA  CPU Microcode Rev # 2.  CONSOLE FW REV# 4.6
                      Standard Microcode Patch    Patch Rev # 1.

       UNIT _VAX1$MKA500:, VOLUME LABEL "BACKUP"

       199637. QIO OPERATIONS THIS UNIT, 0. ERRORS THIS UNIT
       199602. QIO OPERATIONS THIS VOLUME, 0. ERRORS THIS VOLUME
 ******************************* ENTRY    9297. *******************************
 ERROR SEQUENCE 345.                             LOGGED ON:        SID 13000202
 DATE/TIME  3-MAR-2017 22:56:09.54                            SYS_TYPE 01460701
 SYSTEM UPTIME: 1 DAYS 11:41:35
 SCS NODE: VAX1                                                VAX/VMS V6.2

 TIME STAMP  KA690-AA  CPU Microcode Rev # 2.  CONSOLE FW REV# 4.6
                      Standard Microcode Patch    Patch Rev # 1.
 ******************************* ENTRY    9298. *******************************
 ERROR SEQUENCE 346.                             LOGGED ON:        SID 13000202
 DATE/TIME  3-MAR-2017 23:00:19.79                            SYS_TYPE 01460701
 SYSTEM UPTIME: 1 DAYS 11:45:45
 SCS NODE: VAX1                                                VAX/VMS V6.2

 MOUNT VOLUME  KA690-AA  CPU Microcode Rev # 2.  CONSOLE FW REV# 4.6
                      Standard Microcode Patch    Patch Rev # 1.

       UNIT _VAX1$MKA500:, VOLUME LABEL "BACKUP"

       199674. QIO OPERATIONS THIS UNIT, 0. ERRORS THIS UNIT
 ******************************* ENTRY    9299. *******************************
 ERROR SEQUENCE 371.                             LOGGED ON:        SID 13000202
 DATE/TIME  4-MAR-2017 03:06:09.89                            SYS_TYPE 01460701
 SYSTEM UPTIME: 1 DAYS 15:51:35
 SCS NODE: VAX1                                                VAX/VMS V6.2

 TIME STAMP  KA690-AA  CPU Microcode Rev # 2.  CONSOLE FW REV# 4.6
                      Standard Microcode Patch    Patch Rev # 1.
 ******************************* ENTRY    9300. *******************************
 ERROR SEQUENCE 372.                             LOGGED ON:        SID 13000202
 DATE/TIME  4-MAR-2017 03:09:22.47                            SYS_TYPE 01460701
 SYSTEM UPTIME: 1 DAYS 15:54:48
 SCS NODE: VAX1                                                VAX/VMS V6.2

 DISMOUNT VOLUME  KA690-AA  CPU Microcode Rev # 2.  CONSOLE FW REV# 4.6
                      Standard Microcode Patch    Patch Rev # 1.

       UNIT _VAX1$MKA500:, VOLUME LABEL "BACKUP"

       399515. QIO OPERATIONS THIS UNIT, 0. ERRORS THIS UNIT
       199841. QIO OPERATIONS THIS VOLUME, 0. ERRORS THIS VOLUME
 ******************************* ENTRY    9301. *******************************
 ERROR SEQUENCE 728.                             LOGGED ON:        SID 13000202
 DATE/TIME  6-MAR-2017 14:26:12.91                            SYS_TYPE 01460701
 SYSTEM UPTIME: 4 DAYS 03:11:39
 SCS NODE: VAX1                                                VAX/VMS V6.2

 TIME STAMP  KA690-AA  CPU Microcode Rev # 2.  CONSOLE FW REV# 4.6
                      Standard Microcode Patch    Patch Rev # 1.


>  If your tapes are connected that way, directly, cumulative cable length might be an issue.

There are two connectors on the module. Each one goes to a tape drive with about a 6' cable and each drive is terminated. There is another KZQSA port on the far left by the DSSI connector, but that is unused and has an orange cap over it. Here's a picture of it.

IMG_2899.JPG


The drive itself is a DEC drive that was rebuilt and put into a Compaq case so it's externally mounted, which is why there's a difference.

> So I'm kinda curious what type connector(s) are on the back of the box containing this TZ88?  Are they the really old style SCSI 50 pin centronics type with bail locks (about 2" or so) or a smaller, denser connector?

Yup, the old-style 50-pin connector is on the outside. There are two, stacked vertically, with the cable in one and the terminator in the other. No adapters.

Chris Scheers:

Thanks for your reply.

It could be I need a backup tape from a newer OS. I do have a backup tape I made on 5.5.2H4 on 8mm which used to boot (the compression card on that drive failed so the tape doesn't work anymore), but no luck with the 6.2 package. I wonder if I could build a 7.3 kit somehow, such as through the hobbyist program, and boot that.