1. UPDATE: Looks like one of our service providers, Level 3, momentarily blinked off. Out of our hands, but everything’s good now!

    Hey y’all — we forgot to feed Pickles this morning, so he got a little rambunctious in our server closet. As a result, we’re going to be down for a little bit. 

    Our interns are going to get enough bamboo to sate Pickles’ rumbly belly and our systems guys are working as hard as they can to fix the mauled servers. Hang with us, we’ll be back soon!


  2. Earlier today Grooveshark had some problems that you might have experienced while listening to music. Our very own Colin Hostert was nice enough to sit down and explain a bit about what happened.

    We use two pieces of software on each of our webservers. One does all the hard work and runs all the backend code (Apache), the other efficiently servers up static content and caches code that rarely changes (Nginx).

    The caching software has a set amount of space to store all the data it needs to serve up the cached content. Up until today 1GB was more than enough to keep things humming and we couldnt imagine ever filling it up. Now we can imagine filling it up easily. We are still trying to figure out what exactly caused it to suddenly fill up, but we increased it to four times the original size, and decreased how long we cache stuff for. In addition we are taking every step possible so that the machines will email us and text us threatening things if this ever starts to get full again.

    Trust us, we know how much not having your playlists for a few hours can suck—we suffer along with you, since we don’t have any music to play while working hard to fix it!  Sorry again about the hiccup, and serious thanks to Nate, Ed , Myke, and Travis for the help in getting it all fixed today.

     tl;dr -  Cache filled up, wont happen again.