High disk watermark exceeded on one or more nodes, rerouting shards

So, I opened an issue there and here is their reply:

They say, it's an Elasticsearch issue.