The calculated node size_in_bytes value is very different from the actual df value

[524186aa (mingqiang)]

The calculated node size_in_bytes value is very different from the actual df value
I use "size_in_bytes" to check the usage of 2.6tb

The usage of du is 3.7 TB
What's extra? It will be released when it reboots

[524186aa (mingqiang)]

[524186aa (mingqiang)]

help

[TimV (Tim Vernum)]

Please don't post screenshots of text (logs, terminals, etc). Images are hard to read on mobile devices, difficult to search and may be inaccessible to community members with vision impairments.
Please take the time to copy and paste the relevant text instead.

[524186aa (mingqiang)]

Okay, let me correct my description
"roles": [ -
"data",
"ingest",
"ml",
"remote_cluster_client"
],
"attributes": { -
"ml.machine_memory": "269942247424",
"ml.max_open_jobs": "512",
"xpack.installed": "true",
"ml.max_jvm_size": "32212254720",
"transform.node": "false"
},
"indices": { -
"docs": { -
"count": 223613139,
"deleted": 61494359
},
"shard_stats": { -
"total_count": 40
},
"store": { -
"size": "2.1tb",
"size_in_bytes": 2383964133115,
"total_data_set_size": "2.1tb",
"total_data_set_size_in_bytes": 2383964133115,
"reserved": "0b",
"reserved_in_bytes": 0
}
But from the perspective of the operating system, the actual occupancy is much larger
du -sh *
8.2M config
3.7T data
4.0K es.pid
2.8G logs
4.0K version