I85PREP
⌘ K
Dell PowerProtect DD Deploy 2023 study materials
Dell PowerProtect DD Deploy 2023 study resources

Dell PowerProtect DD Deploy 2023: Replication and Disaster Recovery Setup

Dell PowerProtect DD Deploy 2023 · Work through your exam, one topic at a time.

All guide sections

Replication and Disaster Recovery Setup

What this domain covers

Replication questions test topology choices, the exact command sequence, capacity and bandwidth planning, and whether you know what state the destination is in while replication is active.

Replication types

  • MTree replication - replicates one MTree to a peer; the most common choice for granular DR and multi-tenant fan-out.
  • Collection replication - replicates the whole /data/col1 collection, useful for full-system DR where you want one destination mirroring everything.
  • Managed file replication (MFR) - DD Boost feature that replicates specific files with their catalog metadata. Replication is asynchronous, source-authoritative, and included with DDOS at no additional licence.

Configuration sequence

  1. Confirm network reachability and the required ports between peers, and that the destination has enough usable and licensed capacity.
  2. replication add source mtree://<source>/data/col1/<name> destination mtree://<dest>/data/col1/<name>.
  3. replication initialize <context> to seed the baseline - this is a full copy and must be scheduled in a window.
  4. replication enable, then monitor with replication status and replication show stats.
  5. Apply replication throttle for business hours if the link is shared.
  • The destination copy is read-only while replication is enabled. To use it for a restore, break replication, make it writable, then resynchronise or re-initialise afterwards.
  • The destination must run the same or a later DDOS version than the source, and encryption settings must match on both systems.
  • Low-bandwidth optimisation helps on high-latency links; topologies include one-to-one, one-to-many, many-to-one, cascaded and bi-directional (which needs distinct MTree names per direction).

Worked example

A 4 TB daily change must replicate within a 6-hour window: 4 TB / 6 h = 667 GB/hr = 667 x 8 / 3600 = about 1.5 Gbps sustained. A single 1 GbE link cannot meet that; the design needs a 10 GbE path, a throttle that still allows the window, or a shorter retention chain on the source so less data changes. Note the first seed is the full MTree, not the daily change.

Traps that cost marks

  • Describing replication as synchronous or claiming a zero RPO.
  • Forgetting replication initialize, then wondering why the destination has no baseline.
  • Writing to a destination MTree while replication is still enabled - it is read-only.
  • Mismatched encryption settings or an older DDOS on the destination, which blocks the pairing.
  • Bi-directional designs using the same MTree name in both directions.