Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.periphs.scsi > #147

Re: Anybody with SCSI Know How still reading this newsgroup ?

Show all headers | View raw

Peter Dassow wrote:
> 
> [...]
> I've an external HP LTO-2 streamer which works well with a LSI 20160 
> controller, and also with an Adaptec 19160.
> But I've also a Christie LTO-3 streamer which is NOT detected at boot 
> time from the controller, *regardless* if the SCSI bus is terminated or 
> not terminated. Btw. the adaptec controller reports even if a device is 
> properly bus terminated, so you can't do anything wrong.
> 
> When I booted up with the connected LTO-2 streamer, but then exchanging 
> the streamer with the LTO-3 one, I was able to use/control the LTO-3 
> streamer, and also to work with LTO-3 tapes.

I assume that this only works if both drives are configured to the same
SCSI ID.
 
> But exchanging the streamer device everytime AFTER I booted the PC, 
> just to be able to use the external device would not make sense.

Yes. And because the formerly negotated transfer agreement must be
supported, this has worked only by coincidence. In the other direction
(replacing a new drive with an older one), chances are higher that this
will not be the case.

But nevertheless it was a good method to verify that the interface of
the LTO3 drive is not damaged in a way that prevents command execution.

> What can be the reason that the LTO-3 streamer is NOT detected by the 
> SCSI controller, but can be used when booted with the LTO-2 streamer 
> and then be exchanged with the LTO-3 streamer ?

I have not seen this problem for decades. In the early days there were
SCSI devices with a command set that lacks INQUIRY. This is the command
every modern host uses to query information about a SCSI device.
A short time later this command was defined as a mandatory command for
every SCSI device. Is it possible that your LTO3 drive uses a custom
firmware which intentionally lacks the INQUIRY command?
(Maybe the vendor explicitly wants that the drives only work with their
equipment)

Another less likely problem may be the time a target device requires to
become ready after power-on or a bus reset. For power-on the SPI5
document recommends 10s, for reset 250ms. Many hosts will timeout before
these values to speed up the device scan.

Back to comp.periphs.scsi | Previous | Next — Previous in thread | Next in thread | Find similar

Thread

Anybody with SCSI Know How still reading this newsgroup ? Peter Dassow <z80eu@arcor.de> - 2017-01-27 21:49 +0100
  Re: Anybody with SCSI Know How still reading this newsgroup ? Michael Bäuerle <michael.baeuerle@gmx.net> - 2017-01-28 10:21 +0000
    Re: Anybody with SCSI Know How still reading this newsgroup ? Peter Dassow <z80eu@arcor.de> - 2017-02-02 00:42 +0100
      Re: Anybody with SCSI Know How still reading this newsgroup ? Michael Bäuerle <michael.baeuerle@stz-e.de> - 2017-02-02 11:07 +0100
        Re: Anybody with SCSI Know How still reading this newsgroup ? Peter Dassow <z80eu@arcor.de> - 2017-02-06 20:07 +0100
          Re: Anybody with SCSI Know How still reading this newsgroup ? jeffj@panix.com (Jeff Jonas) - 2017-02-12 23:16 -0500

csiph-web