When you get started as a SysAdmin you typically fall into one of 10 categories. You either had a mentor to show you the ropes, or you dredged through it on your on. It seems more and more the group that had or has a mentor is shrinking. I never had a mentor for the things I was most interested in, but my first boss did give good advice, and his words seem to ring in my head often.
Just a little tip, if your looking for bind stats and your chrooted will give you All you have to do is Now after you run rndc stats your named.stats fill will be populated with information.
Had a slightly vexing issue the other day. During a migration the destination host ended up caching some stale DNS entries. I tried a quick restart of nscd to no avail, rebooted the VM, again no joy. Even overriding the host in /etc/hosts wasn’t working. Took a few minutes of digging but what did work was . Just something to remember if you ever seem to have a stubborn caching issue, rebooting/restarting wont always solve the issue.
IMHO one of the great benefits of virtualization is the ability to properly size your guests. Many times 512M memory and an 8G / is plenty. Increasing memory for a virtual machine is typically pretty straight forward, but there are several options when adding disk space. Mounting the space in the file system and doing something with LVM are likely the most common paths. Depending on the situation I typically lean toward keeping my small 8G root and mounting extra storage where needed in the file system.
No one likes to forget to renew a domain. Its not wholly uncommon though. Microsoft forgot to renew passport.com in 1999, and hotmail.co.uk in 2003. Vivendi Universal forgot to renew its MP3.com in 2003, the Washington Post forgot to renew in 2004, and Foursquare forgot to renew their domain in 2010. I stumbled on the check_domain nagios plugin this weekend so I figured I would make some quick modifications and roll it into a Zenpack.
Nginx is a great little web server. I have posted previously about using it as a reverse proxy. Weather your using it as a reverse proxy or as a normal webserver you will probably eventually want to know what its doing over time so you can adjust resources as necessary. The other day I threw together a Zenpack to make it easier to setup monitoring on new Zenoss instances. It contains a command data source “check_nginx_ng” which is a slightly modified version of Chris Kellys check_nginx_ng which in turn was based on check_nginx by Mike Adolphs.
Why do applications have such horrible error messages. Non-specific errors are really not any more helpful than not logging at all. I was recently setting up autofs for mounting home directories from an nfs server. The little buggers refused to work right. All I was seeing in the logfile was a notice that an attempt was made, and failed.
Redundant power is a necessity for any highly available system. Most servers have redundant power supplies and the common design pattern is to have each power supply plugged into a power distribution units that are on separate circuits. One challenge with this type of dealing with this type of design is monitoring the power load. Monitoring A/B power is not as easy as monitoring the individual PDUs. Some servers will draw power from both power supplies, other will draw from one or the other.
Today I woke up to another abrupt Google change, the background image. I know some people want so much candy they can’t see the cane. Some of these changes Google has been making lately makes me wonder if the next time I open up vanilla ice cream I will be looking at a tub of Rocky Road. Yes, the background is only there for a day as publicity for the new feature to make Google more bingesque but I would be pretty upset if my vanilla ice cream turned to Rocky Road as well, even if only for a day.
I don’t have much to say about the loss. Henry Sieff was a brilliant network, security engineer and friend.