Update documentation for "Running a node" + "Troubleshooting"

[0xfdbu]

Describe the issue 描述问题

https://doc.confluxnetwork.org/docs/general/run-a-node/#requirements

While the Linux XFS file system offers performance benefits, it is not as stable; therefore, ext4 is recommended for its reliability and stability.

This part is mentioned, while ext4 do not support copy on write feature which was causing me trouble in this issue

Optional: Affected page(s) or section(s) 受影响的页面

https://doc.confluxnetwork.org/docs/general/run-a-node/#requirements

Optional: Suggested solution 如何解决

1. If the user wants to proceed with ext4 he must make sure that he uses use_isolated_db_for_mpt_table=true in hydra.toml

2. Recommend the user to use XFS file system with Copy on write / reflink feature activated.

Optional: Additional context 问题上下文信息

Here are some logs causing some of the trouble:

INFO Background Snapshott cfx_storage: - COW copy failed, check file system support. Command "cp" "-R" "--reflink=always" "./blockchain_data/storage_db/snapshot/sqlite_f2fd62a797c9ab8a6b0e55db149c55ab8615461.....

[Pana]

So which file system are you using now?
This log COW copy failed is just a info message, you can ignore it.

[0xfdbu]

So which file system are you using now? This log COW copy failed is just a info message, you can ignore it.

after switching to XFS and adding use_isolated_db_for_mpt_table=true , node was fully synced in 6 hours approx