it's recovering, but it took about 2 hours

curl -X GET "10.18.13.64:9200/_cat/recovery/events-training_stats_scalar-d1bd92a3b039400cbafc60a7a5b1e52b?pretty"
events-training_stats_scalar-d1bd92a3b039400cbafc60a7a5b1e52b 0 2.3h existing_store translog n/a n/a 10.18.13.64 clearml n/a n/a 0 0 100.0% 315 0 0 100.0% 93171899010 1544648 1261516 81.7%

Yes, with this size of index on a single shard, it's quite a heavy load on a single-node ES \

Hi @<1578193373062238208:profile|MammothShark25> , this index contains all the scalars reports from experiments. I would assume it is too large for the resources in your system and ES can't load it? Can you attach some ES logs?