Table of contents

  1. Sample scsi-probe-all output
  2. Note about 32-vs-64-bit mode
  3. Note about Transputer spares
  4. Note about Authorize file

Sample scsi-probe-all output

ok probe-scsi-all
/pci@1f,2000/pci@1/SUNW,isptwo@4
Target 0 
  Unit 0   Disk     QUANTUM ATLAS_V_18_SCA  0230
Target 1 
  Unit 0   Disk     QUANTUM ATLAS IV 9 SCA  0909
Target 2 
  Unit 0   Disk     QUANTUM ATLAS IV 9 SCA  0909

/pci@1f,4000/scsi@2
Target 0 
  Unit 0   Disk     QUANTUM ATLAS10K2-TY367JDDD6
Target 1 
  Unit 0   Disk     QUANTUM ATLAS10K2-TY367JDDD6
Target 2 
  Unit 0   Disk     MAXTOR  ATLAS10K4_73SCA DFF0
Target 3 
  Unit 0   Disk     QUANTUM ATLAS10K2-TY367JDDD6

/pci@1f,4000/scsi@3,1
Target 4 
  Unit 0   Processor     TRANST  MatchBox        1.03
  Unit 1   Processor     TRANST  MatchBox        1.03
  Unit 2   Processor     TRANST  MatchBox        1.03
  Unit 3   Processor     TRANST  MatchBox        1.03
  Unit 4   Processor     TRANST  MatchBox        1.03
  Unit 5   Processor     TRANST  MatchBox        1.03
  Unit 6   Processor     TRANST  MatchBox        1.03
  Unit 7   Processor     TRANST  MatchBox        1.03

/pci@1f,4000/scsi@3
Target 0 
  Unit 0   Disk     IBM     DDRS39130SUN9.0GS98E
Target 6 
  Unit 0   Removable Read Only device    TOSHIBA XM6201TASUN32XCD1103

ok 

Note about 32-vs-64-bit mode

Note from October 20, 2008 email re 64-bit:

Just so you know, the 64 bit was caused by me going into single-user mode and transitioning into multi-user mode after that. I believe it did that to me before and I forgot about. So rebooting the box straight into multi-user mode fixes the 64-bit problem.

Note about Transputer spares

As per November 5, 2008 A. Conrad/D. Chan email exchange:

The root transputer board (DAQ17) that was swapped in during June 2008 was labelled with a note saying it had been tested with NIRC2 on September 2, 1999.

There exists one more spare DAQ17 for NIRC2. This one is labelled with a note that says it was tested with NIRC2 on September 9, 1999. It was assigned serial number 1000 on November 3, 2008.

There is also one spare DAQ17 for NIRSPEC.

Note about Authorize file

From A. Honey's August 27, 2010 email:

During testing today with alad server version 4.11 (in response to the problems encountered by A. Conrad during observing the previous night (see K2-19106 and K2-19107), I modified /kroot/kss/nirc2/alad/keyword/Authorize file. That file has entries in it of the form nirc2:nirc2.caltech.edu:1. I modified the file to contain all the references to accounts for observing (user accounts and such) at Keck, and removed the caltech entries. I did that because the problems on the previous night indicated issues with the processing of the Authorize file (seg faults occurred on the response to parsing the file). The modifications I made to the file did not resolve the problem (show and modify failing). I subsequently modified the processing of the Authorize file on the keyword side (show and modify) so that it was effectively neutered. Those changes are not in effect operationally but are in version 4.12.

Later in the day the operational system could not be made functional – the alad_server_bin would repeatedly fail. It turns out that the low level startup code uses the Authorize file to determine if the Transputer boot file is allowed to be downloaded. As I had left my modification in place and deleted the Caltech entries, the system failed. I still do not understand why those original entries in the Authorize file are essential to booting! I will try to figure it out.