Networking, Storage, and Data Protection Services
This section covers Networking, Storage, and Data Protection Services: AHV virtual networking, VLANs and bonds, storage pools and containers, replication factors, erasure coding, snapshots, protection domains, and replication for DR.
Networking and storage rules
- AHV uses virtual switches with uplinks and bonds. Common bond modes include active-backup and LACP; the mode must match the physical switch configuration.
- VLANs segment guest and storage traffic. Verify trunking and allowed VLANs on every uplink.
- A storage pool is the physical capacity layer; a container is a logical storage construct presented to the hypervisor.
- RF2 keeps two copies; RF3 keeps three. Erasure coding can reduce capacity overhead but is not a substitute for every workload's performance and resiliency needs.
- Snapshots are point-in-time copies; protection domains group VMs and schedules for local snapshots and remote replication.
Worked example
A VM disk lives in a container with RF2. A protection domain snapshots the VM every hour and replicates to a remote cluster. The local snapshot protects against logical corruption; the remote copy protects against site loss. If the source site is lost, the recovery plan uses the replicated snapshot on the remote cluster. The container's RF protects against node or disk failure, not against a deleted VM, so both mechanisms are needed.
Traps that cost marks
Do not confuse RF with backups. Do not enable LACP without matching switch configuration. Do not assume a VLAN exists on the trunk just because it exists in Prism. Do not treat erasure coding as always better than RF for every workload.