hi, guys
failed to process cluster event (put-mapping) within 30s problem often happened in my cluster.
My es cluster was used for log management,  lots of index created everyday and I use dynamic mapping adjust to doc indexed.
My cluster config is:
  es version : 5.4.3.  
  master node  num:3
  datenode num: 50
   index num:  8000
  shard num: 20000
When mapping update timeout, the index speed decreases a lot.
So How to solve or debug this problem? Any suggestions?