Pick your current CUCM version, target version, and size. Get a tailored week-by-week migration plan: prep, cutover, rollback — including IM&P, Unity, and UCCX add-ons if you have them.
CUCM migrations get into trouble at version boundaries. 10.5 can't go directly to 15. 11.5 can, but only through PCD. Fresh installs sometimes beat in-place upgrades for old versions because the database schema has drifted. IM&P has to match CUCM version. Unity Connection has its own upgrade path. This tool takes your current version, target version, and deployment size, and produces the specific migration plan that applies — including whether you need an intermediate hop, whether to use PCD (Prime Collaboration Deployment), and what the rollback criteria are.
Pick versions + size, click Generate plan.
Compatibility checks, backups, trial restore, ticket creation, stakeholder comms.
Sequenced upgrade/migration with validation checkpoints.
Monitoring, license re-host, training, documentation refresh.
Exact criteria and steps if you need to abort.
In-place refresh upgrade is the fastest on paper — click upgrade, reboot, done — and is fine when the source version is close to target (12.5 → 15, 14 → 15). It carries every configuration change, hack, and cruft from the current cluster forward. If your cluster has been running since 2016 and was upgraded from 10 to 11 to 12 already, in-place will carry that history forward, including database schema quirks that later bite during operations. Use in-place for clusters that are well-maintained, recently restored from scratch, or have one or two version hops behind.
PCD (Prime Collaboration Deployment) migration is Cisco's tool for taking an old cluster's data and restoring it onto freshly-installed VMs of the new version. It's the right default for most mid-market migrations because it cleans up the OS/platform layer but preserves the CUCM config. PCD is slower than in-place but much safer for big version jumps (9 → 15, 10 → 14). Budget a day per node.
Fresh install + re-provision is the slowest but sometimes the correct choice. If your cluster has legacy dial plan debt, stale CSS/Partition names, per-user Phone Button Templates, and 10 years of "we'll clean that up later", a fresh install is the forcing function to clean it up. Export users, phones, and key configs via BAT, fresh-install the new cluster, import cleanly. This is also the right path if you're consolidating multiple clusters. Budget 2–4 weeks for a 1,000-user cluster depending on how much cleanup you do.
The "auto" recommendation in this tool applies this heuristic: direct hop available + recent maintenance → in-place; multi-hop or older than 3 years → PCD; legacy technical debt or cluster consolidation → fresh install recommended (requires a project, not a cutover weekend).
No — in-place requires an intermediate hop through 12.5. PCD migration can go directly if source is 10.5 SU6+.
Yes — IM&P version must equal CUCM version. Upgrade IM&P immediately after CUCM publisher + first subscriber.
In-place refresh: ~90 minutes per node, serial. PCD migration: 6–8 hours per node, serial. Plan a weekend for any cluster with > 2 nodes.
In-place: the platform keeps an "Inactive Partition" — one command switches back, so the rollback window is long. PCD: same, old VMs remain powered off but intact for up to 30 days. Fresh install: no rollback — validate heavily before cutover.
Only if moving to a major version (e.g., 11 → 14 = new license type required via Cisco Smart Software Manager). Minor upgrades reuse existing.
Recent cluster, well-maintained
Old cluster, hasn't been touched in 5 years
M&A, 4,000 users total
Our UC Modernization Sprint runs your CUCM 9/10/11 → 15 migration end-to-end — plan, cutover, validation, training. Fixed-fee, fixed-timeline.