Long startup time for FNs & BNs

[aWN4Y25pa2EK]

Celestia Node version

v0.21.5

OS

Alpine Linux v3.20

Install tools

docker

Others

No response

Steps to reproduce it

Major delays during the initialisation process while deploying a DA node (FN or BN) over a kubernetes deployment.

Expected result

Normal initialisation expected thresholds ~2 Minutes

Actual result

Actual initialisation thresholds ~30 Minutes on PVCs using 5k iops storage-class and ~10 minutes with 15k iops (measurements taken from lunar)

Relevant log output

Is the node "stuck"? Has it stopped syncing?

No response

Notes

This issue has been observed across multiple clusters. Although it appears to be related to infrastructure storage controller IOPS performance, it would be valuable to understand the root cause of the delay. There must be intensive I/O operations occurring during the file system indexing process.