I don't think it's due to a bug in Lucene, but rather a hardware or OS/network error writing data. Has it happened again? If it's a one-off then this is just elasticsearch detecting a problem with the data and repairing it, but if it's recurrent then we'll need to dig deeper.
1 Like