collabcowboy.net
Expressway-C and Expressway-E cluster operations
🗺️ Atlasa Collab Cowboy tool
Run a staged Expressway upgrade across your C and E pairs. Ordered, verified, recoverable.
Expressway upgrades are a change-window gamble: four node web UIs, a manual order to get right, and no safety net if a peer doesn’t come back. Atlas is lifecycle management for Cisco Expressway-C and Expressway-E clusters: a live status board, cluster-wide maintenance mode, encrypted backups, TAC diagnostic bundles, and a staged, ordered upgrade engine built for multi-peer C and E clusters in DC/DR pairs.
Atlas auto-detects your C/E pairs by reading each C cluster’s traversal-client zone, then drives upgrades in waves you arrange by dragging peers. Per peer it runs maintenance, drains calls, uploads and installs the image, reboots to the new version, and verifies, halting the whole run on any failure, and resuming a halted one where it stopped.
Status board · staged upgrade● verifying
homecore + homeedge · wave 2 of 3 · halts on any failure
| Peer | Role | Status |
|---|
| core1 | Expressway-C | X15.0 · verified |
| edge1 | Expressway-E | X15.0 · verified |
| core2 | Expressway-C | upgrading · reboot |
| edge2 | Expressway-E | X14.3 · queued |
Per peer: Maintenance Drain Upload Install Reboot Verify
Status, Maintenance & Backups
The day-to-day board
A live board shows every peer’s reachability, version, maintenance state, active call count, and replication and alarm health, scraped live and cached so it paints instantly. Set maintenance mode pair-wide to drain a peer, and take encrypted per-peer backups you can download.
- Live status board: reachability, version, calls, replication, alarms per peer
- Auto-detect pairs: reads the traversal-client zone to group and role C/E
- Maintenance mode: set or clear pair-wide, so the cluster routes around a peer
- Encrypted backups: per-peer, tagged and downloadable
Staged Upgrade (Upgrade Trail)
Ordered waves, halts on failure
Upload the image (verified by SHA-512), then arrange peers into upgrade waves by dragging. Peers in a wave upgrade together but only across different clusters, with a pause you control between waves. A halted or interrupted run resumes where it stopped.
- Wave ordering: drag peers; same wave upgrades together, across clusters
- Pause control: manual, timed, or auto-verify between waves
- Per-peer sequence: maintenance, drain, upload, install, reboot, verify
- Recoverable: resume a halted run; peers already on target are skipped
Why teams reach for it
The Expressway change window, de-risked.
One board, not four web UIs. Every C and E peer’s status lives in one place, scraped live, so you’re not tabbing between node consoles to see the cluster.
Pairs detect themselves. Atlas reads the traversal-client zone to find each C’s E partner and role them, instead of you mapping it by hand.
Upgrades run in order. Waves upgrade peers together only across different clusters, so a pair never loses both halves at once.
It stops when something’s wrong. Any failed step halts the whole campaign rather than marching on and taking the edge down.
A halted run resumes. Recovery picks up where it stopped and skips peers already on the target version, so you can finish a half-done cluster.
Backup and diagnostics built in. Take an encrypted per-peer backup before you start, and collect a TAC-ready bundle when you need Cisco.
For C/E pairs in DC/DR you can’t afford to drop
Expressway is the edge for MRA and B2B calling, and a botched upgrade takes remote workers offline. Atlas turns the upgrade into a planned, ordered campaign with maintenance mode, a pre-upgrade backup, and a halt-on-failure engine. So the change window is a runbook Atlas executes, not a manual gamble across four boxes.
Setup & Fit
Works with
Cisco Expressway-C and Expressway-E clusters, including multi-peer DC/DR pairs.
Turn it on
Add a cluster’s primary peer and credentials, then Auto-detect pairs.
Sized for
Single pair up to a multi-pair campaign spanning several C/E clusters.