Sun Fire[TM] Servers (V280R, V480, V490, V880, V890): System Hangs and Interrupt Generation

Asset ID:	1-71-1382079.1
Update Date:	2011-12-05
Keywords:

Solution Type Technical Instruction Sure

Solution 1382079.1 : Sun Fire[TM] Servers (V280R, V480, V490, V880, V890): System Hangs and Interrupt Generation

Applies to:

Sun Fire 280R Server - Version: Not Applicable and later   [Release: N/A and later ]
Sun Fire V480 Server - Version: Not Applicable and later    [Release: N/A and later]
Sun Fire V490 Server - Version: Not Applicable and later    [Release: N/A and later]
Sun Fire V880 Server - Version: Not Applicable and later    [Release: N/A and later]
Sun Fire V880z Visualization Server - Version: Not Applicable and later    [Release: N/A and later]
Information in this document applies to any platform.

Goal

In the event of a hang it is necessary to send an interrupt to the Operating system. This document will describe some of the ways this can be accomplished on the Entry-Level Servers.

Solution

1. Way's to obtain a corefile on soft or hard system hangs on Sun Fire Entry Level Server:

1.1 Ping system and turn Keyswitch to another position (except 'Off/Stby') to make sure system is truly hung.
1.2 Keyboard + Mouse on USB Ports Level 10 Interrupt

1.2.1 STOP-A from Keyboard Level 10 Interrupt
1.2.2 If RSC is configured... rsc> break Level 10 Interrupt

1.3 Serial Port Level 10 Interrupt

1.3.1 Tip Session: '~#' Level 10 Interrupt
1.3.2 Terminal Concentrator: 'ctrl+]'

1.4 Serial Port Level 12 Interrupt

1.4.1 Dumb Terminal/Laptop - Unplug/Plug serial cable Level 12 Interrupt

1.5 XIR Level 15 Interrupt

1.5.1 Solaris 8 with KUP 108528-17 or later; Solaris 9 or later (Required)
1.5.2 RSC Configured (Preferred)... rsc> xir Level 15 Interrupt
1.5.3 Triple-pump - Push the Power button 3 times consecutively, 0.5 seconds apart Level 15 Interrupt (Requires Solaris 8 with 109888-17 or later; or Solaris 9 114126-02 or later)

1.6 Enable Deadman Kernel Add to '/etc/system' 'set snooping=1'

If none of these methods work, then it is necessary to power-cycle to break the hang, then use ASR to disable components may be required to trial and error troubleshoot the hang to a list of possible faulty FRU's causing the hang. Click here for more information on how to determine the type of system hangs and some useful troubleshooting procedures.

2. Hardware Watchdog

Use this method to determine if hardware is wedged. Default behavior of H/W Watchdog is to auto-XIR the system. The intent is for system availability with corefile capture.

Solaris 8 KUP 108528-12 & Platform Drivers 109888-13 (or later) Solaris 9 or later;

Add to '/etc/system' 'set watchdog_enable=1' & reboot system
OBP variables set to 'ok setenv autoboot? true' & 'ok setenv error-reset-recovery? sync'

Formerly Panacea Doc
http://panacea.uk.oracle.com/twiki/bin/view/Products/SystemHangsSunFireV490

References

<NOTE:1008719.1> - Resolving Hardhang problems on Ultra Servers

Attachments

This solution has no attachment