How is disk work handled in your infrastructure…

Depending on the requirements for performance, reliability, and data storage costs, the following approaches are used:

Local Disks (Local Storage): Used for temporary storage, cache, or data that does not require high fault tolerance. They provide high access speed.
Network Attached Storage (NAS): Used for centralized file sharing and backup. Managed as a separate resource on the network.
Storage Area Network (SAN): Preferred for high-performance databases and virtualized environments where low latency and high throughput are required. They represent a separate high-speed network for storage traffic.
Cloud Storage: Used for scalable, reliable, and cost-effective data storage.
- Block Storage: Similar to local disks, mounted as a disk to an instance (e.g., AWS EBS, Google Persistent Disk).
- Object Storage: For unstructured data (files, images, backups). Access via API (e.g., AWS S3, Google Cloud Storage).
- File Storage: Shared access for multiple instances via file protocols (NFS/SMB) (e.g., AWS EFS, Google Filestore).

Standard utilities are used for managing disk spaces and file systems:

# Viewing disk space information
df -h

# Viewing disk partition information
lsblk

# Creating a file system (e.g., XFS)
sudo mkfs.xfs /dev/sdX1

# Mounting a file system
sudo mount /dev/sdX1 /mnt/data

# Adding an entry to /etc/fstab for automatic mounting
// UUID of the file system  mount_point  fs_type  options  dump  pass
UUID=... /mnt/data xfs defaults 0 0

In virtualized environments (e.g., Kubernetes), disk management is done through Persistent Volumes (PV) and Persistent Volume Claims (PVC), abstracting applications from direct storage management. Cloud providers use corresponding integrations (Storage Classes).

Monitoring disk space and input/output performance is carried out using monitoring systems (Prometheus, Nagios) and utilities like iostat, sar.

# Monitoring I/O activity
iostat -xz 1

The choice of specific storage type is determined based on application requirements for performance, volume, availability, and cost. Cloud solutions are preferred for their scalability and manageability unless there are specific requirements for local storage.

How is disk work handled in your infrastructure?