Every time I restart my OS X Mountain Lion (10.9) laptop, it stops using the DNS search domains I've added via the Network preferences pane. I have found that restarting mDNSResponder fixes this issue. There are two ways to do this.

The first is a simple restart by sending a SIGHUP to the process:

$ sudo killall -HUP mDNSResponder

The other option is to stop and start the process with the launchctl command:

$ sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.mDNSResponder.plist
$ sudo launchctl load -w /System/Library/LaunchDaemons/com.apple.mDNSResponder.plist

I tend to clear out large chunks of email all at once. The most common thing I do is clear out all mail older than 30 days.

In mutt do the following:

• Shift-D to delete messages matching a pattern. This will return a prompt that says Delete messages matching:.
• ~d > 30d to delete messages older than 30 days.
• Enjoy the extra disk space you just created.

On my web server, I keep my web sites in /var/lib/www. Each web site has a dedicated logs directory. Since logrotate is already rotating Apache logs on the server, I'd like to continue using that to rotate these log files as well. Below, is a Puppet snippet that uses Augeas to modify the stock Debian logrotate configuration file for Apache to include my custom log directories.

augeas { "apache2/logging/sites":
    lens    => "Logrotate.lns",
    incl    => "/etc/logrotate.d/apache2",
    changes => [
        "ins file after rule[file='/var/log/apache2/*.log']/file[last()]",
        "set rule[file='/var/log/apache2/*.log']/file[last()] '/var/lib/www/*/logs/*_log'",
    ],
    onlyif  => "match rule[file='/var/log/apache2/*.log']/file[.='/var/lib/www/*/logs/*_log'] size == 0",
    require => Noop["apache/installed"],
    before  => Noop["apache/configured"],
}

I stumbled across this problem when a daily cronjob on one of our Debian servers complained about this. A lot. If you run across this on one of your systems, there is a quick fix.

In our case, it turned out that /usr/bin/mandb could not read /etc/passwd because the mode was 0600 rather than 0644. A quick change (chmod 644 /etc/passwd) and things were back to normal.

Thanks to an error report from one of my backup jobs, I noticed that /Users/slap/Library/Logs/DiagnosticReports was filling up with Finder crash reports (files of the type *.crash). These files were being created once a minute. Additionally, icons on my Desktop were blinking once a minute.

Upon investigation, it seems that Google Drive was the culprit. Specifically, Google Drive's ability to update icons to indicate their sync stats. By disabling the "Show file sync status icons and right click menu" checkbox in Google Drive's Preferences window, I was able to stop the crashing.