Dynamic graphing is uber cool
As we continue to get more servers to run our websites, one of the complications that arises is monitoring. As part of the solution, I created a few service monitors to let us know if something goes down. However, being proactive is better than being reactive. Enter graphing. We now have a dedicated machine showing us cpu and memory line graphs with overlaid text for each server, as well as some bar graphs for network activity to each machine. In the interest of going modern and trendy, I even decided to use a CF card for booting and slax as the distribution for this machine. Now, on to monitoring the memcached hits vs. misses.