We periodically see core data left on disk after deletes. For example, I split a shard "shard54" into 8 pieces. They are all active now. And I deleted the original, now inactive, shard via the Solr UI.
It is now a day later and I noticed that the data from the original shard is still on disk. There is no reference to this shard in the UI or in the collection's state.json file. I restarted the Solr instance and it did not clean up the data on disk which remains there. You can see the truncated output of `du -h` on the solr VM below. It still has the original shard data. Is this a known bug? I am on Solr 8.11.3. Thanks, Zack ``` 0 ./userfiles 0 ./filestore 0 ./my_shard54_replica_n1607/data/snapshot_metadata 9.8M ./my_shard54_replica_n1607/data/tlog 29G ./my_shard54_replica_n1607/data/index.20240204025629715 29G ./my_shard54_replica_n1607/data 29G ./my_shard54_replica_n1607 0 ./my_shard54_0_replica_n3145/data/snapshot_metadata 1.2G ./my_shard54_0_replica_n3145/data/index 624K ./my_shard54_0_replica_n3145/data/tlog 1.2G ./my_shard54_0_replica_n3145/data 1.2G ./my_shard54_0_replica_n3145 0 ./my_shard54_1_replica_n3146/data/snapshot_metadata 1.5G ./my_shard54_1_replica_n3146/data/index 652K ./my_shard54_1_replica_n3146/data/tlog 1.5G ./my_shard54_1_replica_n3146/data 1.5G ./my_shard54_1_replica_n3146 0 ./my_shard54_2_replica_n3147/data/snapshot_metadata 2.3G ./my_shard54_2_replica_n3147/data/index 548K ./my_shard54_2_replica_n3147/data/tlog 2.3G ./my_shard54_2_replica_n3147/data 2.3G ./my_shard54_2_replica_n3147 <etc> ```
