home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.os.vms
- Path: sparky!uunet!charon.amdahl.com!pacbell.com!ames!sun-barr!cs.utexas.edu!sdd.hp.com!decwrl!pa.dec.com!engage.pko.dec.com!nntpd.lkg.dec.com!ryn.mro4.dec.com!star.enet.dec.com!parris
- From: parris@star.enet.dec.com (Keith B. Parris)
- Subject: Re: Remotely rebooting vms
- Message-ID: <1993Jan26.144801.25061@ryn.mro4.dec.com>
- Sender: news@ryn.mro4.dec.com (USENET News System)
- Organization: Digital Equipment Corporation
- References: <1993Jan14.114547.279@vulcan.resmel.bhp.com.au> <1993Jan15.172445.18573@nessie.mcc.ac.uk>,<1993Jan15.192425.13304@dbased.nuo.dec.com> <1j8v9eINNng8@gap.caltech.edu>
- Date: 26 JAN 93 09:20:51 EST
- Lines: 76
-
-
- In article <1j8v9eINNng8@gap.caltech.edu>, carl@SOL1.GPS.CALTECH.EDU (Carl J
- Lydick) writes...
- >In article <1993Jan15.192425.13304@dbased.nuo.dec.com>,
- >lionel@quark.enet.dec.com (Steve Lionel) writes:
- ...
- >>But to the point (or points)... First, the major reason VMS doesn't shut down
- >>to the console prompt is that the way to do that is to execute a HALT
- >>instruction and since most people have their systems set up to do a reboot
- >>on halt, that's not very useful.
- >
- >Actually, the options are:
- > 1) Do a reboot on a BUGCHECK; or
- > 2) Don't to a reboot on a BUGCHECK;
-
- You're probably thinking of BUGREBOOT. What Steve was probably talking about
- was the default HALT action setting stored in NVRAM by the console firmware,
- which controls what happens when a HALT instruction is executed in kernel mode
- (unless VMS specifies a different action by writing to another area in NVRAM).
- The default action setting is controlled by:
- >>> SET HALT n !on 3100s and VAXstation 4000s, or
- >>> T 53
- 2 ? >>> n !on 2000s
- The default setting is 2, which means to reboot on a HALT. You can set it
- to 3, which causes an entry to console mode on a HALT, instead of a reboot.
-
- When you specify to SHUTDOWN.COM that an automatic reboot is desired, VMS
- overrides this default action and a reboot occurs. Otherwise, if a HALT
- instruction were to be executed in kernel mode, this setting controls the
- behavior. It would be nice if VMS would allow you to specify during shutdown
- that you want to enter console mode, but today, after printing out the SHUTDOWN
- COMPLETE message, it just enters an infinite loop inside EXCEPTION.EXE, at
- location 4022(hex) for V5.5-2, last I looked.
-
- If you wanted to use the halt action setting to halt in console mode, you'd
- want to execute a HALT instruction after shutdown was complete. The most
- obvious way might be to patch EXCEPTION.EXE, but following is a way using
- supported tools:
-
- Set things up in NCP on the boot node(s) so a reboot will load an image which
- does a HALT instruction. The READ_ADDR.SYS program, which is normally used to
- find out the Ethernet address of a satellite node, just happens to do a HALT
- instruction when it's run, and is readily available, so it will do the job
- nicely.
-
- To HALT the satellite on the next reboot attempt:
-
- $ MCR NCP
- NCP> CLEAR NODE satellite LOAD ASSIST PARAMETER
- NCP> CLEAR NODE satellite LOAD ASSIST AGENT
- NCP> SET NODE satellite LOAD FILE MOM$LOAD:READ_ADDR.SYS
-
- Now, shut the satellite down normally, but with an immediate reboot specified.
-
- $ MCR SYSMAN
- SYSMAN> SET ENVIRONMENT/NODE=satellite
- SYSMAN> DO @SYS$UPDATE:AUTOGEN REBOOT
-
- On reboot of the satellite, the READ_ADDR.SYS program will be loaded, dutifully
- print out the Ethernet address, and the system will "halt" in console mode.
-
- Later, when you're ready to allow the satellite to boot normally, you need
- to fix up the node information in NCP, so that the satellite will load VMS,
- rather than the READ_ADDR.SYS program, when it is booted.
-
- To allow the satellite to reboot normally:
-
- $ MCR NCP
- NCP> CLEAR NODE satellite LOAD FILE
- NCP> SET NODE satellite ALL
- __________---_---_-_-_-_-_____-_----____---__-__-_-_-----_-_--_--___-_-__---__-_
- Keith B. Parris, Digital Equipment Corporation, Alpha VMS Clusters & I/O Group
- ----_-____-_--_-__-_---_-___-_-_-_--__-_-__---__-_---_-___-_-_-_-_____--_--___-_
- Anything blatantly opinionated or which is later found to be inaccurate was
- mine, of course. Digital gets the credit for any actual facts which appear.
- -_-_--__-_-__----___---__-__-_-_-_-_____-__--_____---_-___-_-_-_--__-_--_--___-_
-