Issue has been resolved, all clusters and shards are operational.
We've made a fix and are currently gracefully deploying it to all clusters.
We've identified the error and have resolved it, deploying it now to one of the clusters that had issues.
If all goes well then we'll continue deploying it to the rest of the clusters.
We're currently investigating this issue. This is affecting a handful of servers, we're unsure of how many as of right now though.
