Service

The whole fleet on supported, patched software, upgraded in planned waves

We automate the software upgrades across your network fleet: a readiness check, a backup, the upgrade, and checks that confirm each device is back to normal. The upgrades run in parallel, wave by wave, inside your maintenance windows, and HA pairs fail over so the service stays up.

Measured impact

What changes for your organization

  • Today

    Upgrades that take months

    Each device is upgraded by hand, one window at a time, so a fleet-wide upgrade runs for months and the next release arrives before it ends.

    With fleet upgrades

    The fleet current, on schedule

    Upgrades run in parallel waves inside your maintenance windows, so the whole fleet reaches the target release on a date you plan.

  • Today

    A growing patch backlog

    Devices stay on old or unsupported releases because every upgrade costs engineer time, and known vulnerabilities stay open.

    With fleet upgrades

    Supported, patched software

    Every device runs a supported release, and a security patch reaches the fleet in the next campaign.

  • Today

    Every device a special case

    Standalone units and HA pairs, physical and virtual, several models and releases: each one needs its own procedure and its own expert.

    With fleet upgrades

    One procedure for every variant

    The same guarded sequence handles each kind of device in the fleet, so the result is consistent and any engineer can run it.

  • Today

    Outages during the window

    A failed upgrade is found by users, and the way back is rebuilt by hand in the middle of the night.

    With fleet upgrades

    Service that stays up

    HA pairs fail over during the upgrade, readiness checks stop a risky upgrade before it starts, and a backup is ready for rollback.

  • Today

    Proof pieced together

    Which device runs which release, and whether it came back healthy, is collected from spreadsheets and device logs.

    With fleet upgrades

    A record of every upgrade

    The checks before and after each upgrade are kept, so the state of the fleet and every upgrade’s result are on record.

What we deliver

  • Fleet inventory and target releases

    Every device, its model, role and current release, with the target release for each and the upgrade path that gets it there.

  • Readiness checks

    Before anything changes, the automation confirms the device is ready for the upgrade, and stops it when it is not.

    Read the case study →
  • Upgrade automation

    A library and playbooks that upgrade standalone devices and HA pairs, physical and virtual, with failover so the service stays up.

    Read the case study →
  • Post-upgrade validation

    Pre- and post-validation checks compared around each upgrade, confirming the device is back to its normal state.

  • Backup and rollback

    The configuration backed up automatically before each upgrade, so there is always a known state to return to.

  • Waves in parallel

    Workflows run in parallel and grouped in waves, so each maintenance window carries more upgrades and the fleet moves on a plan.

How it runs

The fleet’s size sets the number of waves; the sequence each device goes through stays the same.

  1. Discovery

    Your fleet, its releases and your upgrade process today, the target releases, and a plan of waves fitted to your maintenance windows.

  2. Automation build

    Readiness checks, backup, upgrade and post-upgrade validation for each kind of device in the fleet, built and tested in a lab.

  3. Pilot wave

    A first, small wave on production devices, with the results reviewed against the checks before the next wave starts.

  4. Fleet waves

    The rest of the fleet upgraded in parallel waves, each inside a maintenance window, with rollback ready for any device.

  5. Handover

    The automation documented and handed to your team for the next release, or run by us as an ongoing service.

Two ways to run it

  • An upgrade campaign

    We build the automation for your fleet and run the upgrade to the target release with your team, then hand the automation over for the next one.

  • Upgrades as an ongoing service

    We build the automation and keep running it, release after release, so the fleet stays on supported software year round.

Proof

Why CodiLime

  • Proven at fleet scale

    Automation we built and run upgrades 10k+ firewalls a year for a tier-1 vendor's Professional Services organization, across airlines, MSPs and banks.

  • Uptime that holds

    For a healthcare enterprise, 500+ firewalls were upgraded in less than 3 months, with HA pairs failing over so the services they protect stayed up.

  • Network engineers who write software

    One team that knows the devices and the protocols, and builds the upgrade automation in Python, Ansible and AWX.

FAQ

Good to know

  • Which devices can you upgrade?

    The automation is written per platform: standalone units and HA pairs, physical appliances and virtual ones. The discovery lists the models and releases in your fleet and what each one needs.

  • How do you keep the service up during an upgrade?

    HA pairs fail over while one member upgrades, readiness checks stop an upgrade on a device that is not ready, and every upgrade runs inside your maintenance window.

  • What happens when an upgrade fails?

    The post-upgrade checks show it at once, and the configuration backup taken before the upgrade is the way back, so the device can be rolled back quickly.

  • How long does a fleet upgrade take?

    It depends on the size of the fleet and the number of maintenance windows you have. The discovery ends with a plan of waves and a date for the last one.

  • Who runs the next upgrade?

    Your team, with the automation documented and handed over, or CodiLime, running the upgrades as an ongoing service release after release.

Start with an upgrade assessment

Tell us about your fleet. We map the devices, their releases and your maintenance windows, design the upgrade automation, and hand you a plan of waves that brings the whole fleet to a supported release.