Questions and Answers

Thanks to Robert Banniza for a correction to the February 2003 (http://www.samag.com/documents/s=9023/sam0402h/0402h.htm) Q&A about booting a Sun V240 from Solaris 8. Although Sun specifies on their Web site that the V240 should boot from Solaris 8 02/02 , Solaris 8 05/03 is actually required because the machine will issue a kernel panic due to missing drivers.

QSun specifies on their Web site that the V240 should boot from Solaris 8 02/02 , Solaris 8 05/03 is actually required because the machine will issue a kernel panic due to missing drivers.Q Our group just inherited a Sun AXi that someone had been using as a desktop machine. We were going to turn this box into a small server, but it has a very annoying problem. Every time the machine resets, it won't auto-boot. Instead, it gives the error trap 3e. If you type in boot from the console after the error, the machine comes up just fine.

I've tried everything I can think of to diagnose/fix this problem. I increased the diagnostic level, and there doesn't appear to be anything really wrong. I tried enabling diag-switch? and setting the diag-device to disk, but it just gives the error twice. I tried reinstalling the OS, reinstalling the boot blocks, and fsck'ing the disk. Nothing ever shows up as actually being a problem, but the machine just fails to auto-boot. Do you have any suggestions?

A It sounds like your boot drive is an IBM. There's a known issue with some IBM drives not being ready after a reset. The workaround is to force a disk probe upon power on so that the disks are recognized and ready to boot from by the time the machine tries to boot. To accomplish this, drop the machine to the ok prompt and use nvedit to add the some lines to the nvramrc. The following is a quick key-binding list for nvedit:

• Control-B: Move backward one character.
• Control-C: Exit the editor and return to the OpenBoot command interpreter. The temporary buffer is preserved but is not written back to NVRAMRC. (Use nvstore afterwards to write back the temporary buffer.)
• Control-F: Move forward one character.
• Control-K: If at the end of a line, join the next line to the current line (that is, delete the newline).
• Control-L: List all lines.
• Control-N: Move to the next line of the NVRAMRC editing buffer.
• Control-O: Insert a new line at the cursor position and stay on the current line.
• Control-P: Move to the previous line of the NVRAMRC editing buffer.
• Delete: Delete the previous character.
• Return: Insert a new line at the cursor position and advance to the next line.

For the entire command-line editor key map, see:

http://docs.sun.com/db/doc/805-4434/6j46vcb5g?a=view

nvramrc=probe-all install-console banner
" Probing all to work around IBM boot disk trap 3e error" type cr
" /pci@1f,0/pci@1/scsi@1" " show-children" execute-device-method drop

nvstore ok
setenv use-nvramrc? true

Q I'm trying to do some performance tuning on a FreeBSD 4.9-STABLE machine, so I've whipped up a Bourne shell script to time some functions. I'm then parsing and correlating the output from the time command to generate some statistics. Because I needed some extra functionality, I switched to bash. Unfortunately, the time command doesn't seem to be outputting what I expect. Here's the simplest case of the two scripts, with the bash script first:

#!/usr/local/bin/bash
time ls -a

.       ..

real    0m0.004s
user    0m0.000s
sys     0m0.003s

#!/bin/sh
time ls -a

.       ..
        0.00 real         0.00 user         0.00 sys

A You're getting different output because bash reserves the time keyword and prints it out in its own format. You can reformat the bash output to appear like /usr/bin/date output would in the Bourne shell by modifying the TIMEFORMAT variable in your script. Here's an excerpt from the BASH FAQ:

The value of TIMEFORMAT is a string with '%' escapes expanded in a fashion similar in spirit to printf(3). The manual page explains the meanings of the escape sequences in the format string.

If TIMEFORMAT is not set, bash acts as if the following assignment had been performed:

TIMEFORMAT=$'\nreal\t%3lR\nuser\t%3lU\nsys\t%3lS'

The POSIX.2 default time format (used by 'time -p command') is

TIMEFORMAT=$'real %2R\nuser %2U\nsys %2S'

The BSD /usr/bin/time format can be emulated with:

TIMEFORMAT=$'\t%1R real\t%1U user\t%1S sys'

The System V /usr/bin/time format can be emulated with:

TIMEFORMAT=$'\nreal\t%1R\nuser\t%1U\nsys\t%1S'

The ksh format can be emulated with:

TIMEFORMAT=$'\nreal\t%2lR\nuser\t%2lU\nsys\t%2lS'

You want the BSD format, so your script would be written as:

#!/usr/local/bin/bash
TIMEFORMAT=$'\t%1R real\t%1U user\t%1S sys'
time ls -a

A Sendmail uses rulesets from the sendmail.cf file to process entries from the access.db file (it's just a lookup table for the rulesets). Therefore, specific parts of a message are processed before others. The ordering of lines in the access file is inconsequential and may not correspond to the order in which the entries are read by the rulesets.

The sendmail rulesets are organized such that, by default, checks on the sender address occur after the check on the connecting machine name/IP. If your LHS specifies a hostname or IP addresses, then your OK will be processed. If your LHS specifies an entire email address (username and machine name), then the mail may be rejected because the DNSBL lookup happens first. Here's an example access file to make that a bit clearer:

that.domain           OK
user@other.domain     OK

To work around this behavior, you can specify delay_checks in your sendmail.mc file and rebuild your sendmail.cf file:

FEATURE('delay_checks')dnl

  FEATURE('delay_checks', 'friend')dnl
  FEATURE('delay_checks', 'hater')dnl

If such an argument is given, the recipient will be looked up in the access map (using the tag Spam:). If the argument is 'friend', then the default behavior is to apply the other rulesets and make a SPAM friend the exception. The rulesets check_mail and check_relay will be skipped only if the recipient address is found and has RHS FRIEND. If the argument is 'hater', then the default behavior is to skip the rulesets check_mail and check_relay and make a SPAM hater the exception. The other two rulesets will be applied only if the recipient address is found and has RHS HATER.

This allows for simple exceptions from the tests, e.g., by activating the friend option and having Spam:abuse@ FRIEND in the access map, mail to abuse@localdomain will get through (where "localdomain" is any domain in class {w}).

A second parameter, n, can be specified with the delay_checks feature to turn off backward compatibility for versions of sendmail prior to 8.12, which used a different syntax:

FEATURE('delay_checks', 'friend', 'n')dnl

or

FEATURE('delay_checks', 'hater', 'n')dnl

A There are a few ways you could go about this, but the easiest way is probably to change the virtual host directive and redirect all traffic for port 80 over to the https instance on port 443. The virtual host directive for port 80 in your httpd.conf would look something like the following:

 <VirtualHost *:80>
  ServerName www.my.domain
  Redirect / https://www.my.domain/
</VirtualHost>

A There's a LDAP patch for Solaris 8, 108993, that includes password aging with pam_ldap. Be sure not to apply version 29 of this patch because it had issues. The latest good revision of this patch is 26, and this is the documenting text from sunsolve.sun.com:

Patch-ID# 108993-26
Keywords: security ldap ldapclient libldap automountd libthread libc sigbus
Synopsis: SunOS 5.8: LDAP2 client, libc, libthread and libnsl libraries patch
Date: Sep/24/2003

******************************************************

Patch 108993-26 was re-instated on Nov 4, 2003

Reason:
Patch 108993-29 has been withdrawn:
After applying patch 108993-29, the init(1M) command will hang
when changing the system run level to 0, 5 or 6
(halt, power off, reboot). In addition, it is not possible to log
|in to the system in single user mode.

Please reference bug 4947516 for further information.

Recommendation:
Backout patch 108993-29 or replace /etc/lib/nss_files.so.1 with an
older version of /etc/lib/nss_files.so.1 for Solaris 8. This must be
from patch revision -26 (or earlier).

|***********************************************

4357827 pam_ldap should fully support password aging

perl -pi.bak -e "s/foo\s(1234.+?)\sbar/baz $1 bop/m;" *.html

foo 1234abcd bar foo 1234wxyz bar

baz 1234abc bop foo 1234wxyz bar

baz  bop foo 1234wxyz bar

A Your regular expression is fine, but your shell is getting in the way. Because you're using double quotes, the $1 is being expanded and substituted by the shell. Because your shell has nothing stored in $1, there's nothing to replace it in your output file. Try using single quotes around your Perl command or create a Perl script instead of calling perl from the command line.

Amy Rich, president of the Boston-based Oceanwave Consulting, Inc. (http://www.oceanwave.com), has been a UNIX systems administrator for more than 10 years. She received a BSCS at Worcester Polytechnic Institute, and can be reached at: qna@oceanwave.com.