Lucene file deletion seems to hang on Windows Server [LUCENE-10053]

Hi everyone,

Firstly, i would like to apologize if it's not the right spot to post my issue, but it's been month me and my team are stuck on this.

We're using an ElasticSearch cluster for months now. Sometimes, a nodes is just getting stuck for a unknown reason.

When i'm trying to check threads contention, i'm getting this stackTrace : 

[Elastic_hot threads](https://pastebin.com/g4mPD0xA)

It seems that the inner file deletion is getting stuck by something .. In my opinion, a deletion is an atomic operation so i can't figured it out how it can hangs forever. I need to restart ElasticSearch to have my node back (i have to kill the process actually). I disabled Antivirus and backup process on these servers and still getting these errors randomly 2 or 3 times per month.

Do you have any clue on what i should care about ?

Thank you for your help

---
Migrated from [LUCENE-10053](https://issues.apache.org/jira/browse/LUCENE-10053) by Rayane, updated Sep 21 2021
Environment:
```
Windows Server 2019 16Go RAM 8vCPU
```

Attachments: [MicrosoftTeams-image.png](https://apache.github.io/lucene-jira-archive/attachments/LUCENE-10053/MicrosoftTeams-image.png)

Start by reviewing the linked hot-threads stack trace and the migrated LUCENE-10053 report, including the Windows Server 2019 environment and attachment. Investigate the reported file-deletion hang and identify a reproducible cause or reliable reproduction; done means documenting the cause and a verified resolution or next diagnostic step.

Written by the indexing model from the issue text.