Skip to main content

​
DP Offline Ubuntu Upgrade Guide

This site is an operator runbook. If this is your first time performing the upgrade, start with Step 1 and continue in order. You should not need to assemble a procedure from separate technical pages.
First time here? Start with 1. Upgrade Work Plan, then follow the numbered pages in the left navigation from top to bottom.

​
The complete flow

The Mirror Server downloads and validates the required files in advance. During the production upgrade, DP hosts retrieve prepared material from that server over HTTP/TCP 80.

​
Follow these five steps

1. Upgrade Work Plan

Confirm the upgrade path, maintenance sequence, snapshots, cluster order, and go/no-go gates before touching any DP.

2. Ubuntu 24.04 Server Installation

Build the Mirror Server when a new server is required, then verify disk, network, and SSH access.

3. Mirror Server Setup

Install Mirror Manager, prepare the files, enable HTTP distribution, and reach Upgrade Readiness = PASS.

4. DP Offline Ubuntu Upgrade

Use fresh Menu 7 commands for every OS hop and for DP 6.6.0 Phase 2 bringup.

5. Post-Upgrade Validation

Verify Ubuntu, DP version, DL/DA readiness, services, NTP/DNS, application health, and recovery-snapshot cleanup criteria.
Do not begin DP-side work until the current Mirror Server configuration reports Upgrade Readiness = PASS and fresh Menu 7 commands have been generated.

​
Overall work plan

The following example shows the complete flow for a 3-node DL cluster + 3-node DA cluster. Smaller deployments follow the same stage order; use the current Menu 7 output for the exact commands and node-specific procedure.
DP Ubuntu 24.04 offline upgrade work plan

Example end-to-end maintenance flow for a 6-node DP deployment

​
Which path applies to my DP?

The current Mirror Manager target is DP 6.6.0. Historical screenshots or recordings may show 6.5.0; do not copy those release-specific commands into the active workflow.

​
Three rules to remember

  1. Prepare first. Complete the Mirror Server work before the production maintenance window whenever possible.
  2. Use fresh generated commands. Menu 7 is the authoritative command source for the current Mirror Server IP, mode, and cluster configuration.
  3. Keep recovery available. Create the required hypervisor snapshot(s), follow any additional powered-off snapshot gate printed by the current upgrade client, and keep recovery points until final validation and change-control closure.

​
If something does not match the guide

Use Troubleshooting for the failed layer, Retry & Recovery for interrupted operations or Phase 1 state, and Video Guides for supplemental visual references. Lower-level implementation material is separated into the Reference tab so the numbered runbook remains the primary operator path.