musings and one liners

How to Improve the Resilience of your MPP Sharding

By now you’ve probably heard of Foursquare’s MongoDB outage, leading to several hours of downtime a couple weeks ago (if not, check the post mortems of Foursquare and MongoDB). The High Scalability blog now takes a look at Sharding in general, explains what it is, defines sharding trouble categories, and finally suggests how to deal with your troubled shards. Excellent article: Troubles with Sharding – What can we learn from the Foursquare Incident?