The Life of a Sysadmin

Just had a repeat of the weird mouse-X disconnect I've encountered before. This time though, I'm running Debian Etch — so no more blaming the problem on SuSE (as I secretly always did :-).

One noticeable problem this time was that the middle button did not work, making click-to-paste impossible; I even ran xev and saw no events for middle-clicking. (This in addition to clicking being inconsistent, the client receiving the click being inconsistent, etc). Running cat /dev/input/mouse0 did not work. What did work was disconnecting the mouse (a USB 3-button optical jobbie), then plugging it back in. Sure, coulda been the mouse driver, or X, or something, but I wonder if the hardware itself — whatever little controller chip is in there — maybe got wedged. Interesting to think about…

Thanks to the good folks who wrote this page, I was able to come up with some quick graphs showing how often our RealMedia files are used:

All it took was five minutes mucking about ... cf OpenOffice.org, where I could not, for the life of me, figure out how to get a decent-looking output.

How to quiet noisy cron entries that send far too much to STDERR:

exec 3>&1 ; /path/to/script 2>&1 >&3 3>&- | egrep -v 'useless|junk' ; exec 3>&-

I've been very busy of late, but the biggest news is that I've started a 3-month temporary part-time assignment here. It's a neat place, and feels a lot like a software startup. Even though it's a small group, they've got certain hardware requirements that are a lot bigger than what I've worked with before; it'll be interesting, to say the least.

• When trying to change an Exchange password using Outlook Web Access at $UNIVERSITY, a user got "Error number: -2147024891". winerror.h describes this as a General access denied error. In the end, it turned out that when the account was created, the "user cannot change password" option was checked. Hope that'll help someone else's google-fu…
• I had managed to miss this quick command-line check for Dan Fucking Kaminsky's DNS bug, aka CERT VU#800113. Just run:

dig +short porttest.dns-oarc.net TXT

and watch the skies.

• http://metasploit.com/dev/trac/changeset/5579
• http://blog.wired.com/27bstroke6/2008/07/dns-exploit-in.html
• http://www.caughq.org/exploits/CAU-EX-2008-0002.txt

Andy just pointed out to me that it's Sysadmin Day, which I'd totally forgot about. So here's to ya, everyone!

Which means I'll have to save the dream I just had (one of the Rohirrim being held hostage by the housewives of the OC at one of their garden parties, and insisting, with murderous glint in their eyes, that he finish all of his cake before leaving despite the rest of the Rohirrim showing up to rescue him, and me trying desperately to resolve the standoff by telling Peter Gallagher that, really, there's still time to let him go) for another entry…/me shakes head.

Ran into a problem today when adding this stanza to cfengine on a Debian Etch machine:

editfiles:
        { /etc/aliases
                AppendIfNoSuchLine "root: sysadmin@pims.math.ca"
                DefineClasses "rebuild_aliases:restart_postfix"
        }

The cfengine reference file I've got, which sez it's for version 2.2.1, says you can define multiple classes in DefineClasses (or DefineInGroup), as long as they're separated by commas, spaces or dots. (The version in Etch is 2.2.20.)

However, when I ran cfagent, it just hung immediately after performing the edit, and gave this error when I ctrl-c'd it:

cfengine: Received signal 2 (SIGKILL) while doing [pre-lock-state]

Running cfengine with -d2 showed endless repetitions of AddClassToHeap() at this point, so either there's something wrong with my syntax or there's a bug in cfengine. (I'm guessing the former.) Searching for pre-lock-state and cfengine only turned up cases where the clients were syncing with the master; thus this note.

The fix was to just make it one class:

                DefineClasses "rebuild_aliases"

Asking to restart Postfix was probably a bit of overkill anyhow...