Restart every room overnight
Reboot the screens and codecs in every meeting room at 06:00, a couple of seconds apart rather than all together. For IT and AV teams whose kit needs a weekly or nightly power cycle to stay reliable.
What this workflow does
- A Schedule trigger fires at 06:00.
- Find Devices collects the outlets that feed the room equipment across the whole building.
- Power Cycle Outlet cuts each outlet for ten seconds and lets it come back, with a two-second gap between one room and the next.
- A branch on the result tells someone if part of the building didn't answer.
The gap is the important part. Two hundred screens powering up in the same instant is a current spike the building has to absorb, and two hundred devices reaching for the network together is its own problem. Because the power unit times the off period itself, spacing the commands spaces the power-on events: at 2s, the rooms come back gradually over about seven minutes without the workflow having to wait for each one.
Prerequisites
- Workflows enabled for your organization. See Enable workflows for your org.
- Gude Expert Power Control units connected, with Permit CLI commands enabled on each one.
- The outlets that feed room equipment identifiable as a group. A tag is the easiest way — tag them once and Find Devices can pick them out from everything else in the rack.
Build it
1. Create the workflow
Workflows → New workflow → name it Restart room equipment → Create.
2. Add the Schedule trigger
Add node → Triggers → Schedule. Set it to run daily at 06:00 in your building's timezone. Weekdays only is usually what you want — nobody needs a Sunday reboot.
3. Find the outlets
Add node → Conditions → Find Devices.
- Space — the building, with recurse switched on.
- Integration — your Gude integration.
- Tag — the tag on the outlets that feed room equipment.
Wire the found output onward. The none output is worth wiring to a notification too: it means the tag matched nothing, which is a silent no-op otherwise.
4. Add the Power Cycle Outlet action
- Add node → Actions → Power Cycle Outlet.
- Outlets — template it from Find Devices:
{{ .findDevices.deviceIds }}. - Off for (seconds) — 10 is enough for most equipment; give a PC longer.
- Stagger → Between commands —
2s.
5. Catch a partial rollout
The action has one output, so add an If after it to do the branching. Give it two rows — the Not attempted list Is Not Empty, and the Unconfirmed list Is Not Empty — set the combiner to Any, and wire the true edge to a Send Direct Message action. Those two buckets mean something specific:
- Not attempted — the rollout ran out of time and never reached those outlets. Nothing was sent to them.
- Unconfirmed — the command went out but the unit never answered. Usually Permit CLI commands switched off, or the unit offline.
Try it
Don't wait for 06:00. Point a copy of the workflow at a single room's outlet, run it from the editor, and watch the screen go off and come back. Then widen the device list.
The Executions panel shows which list each outlet ended up in.
Watch out for
- Check it can finish. The action refuses a rollout that won't fit in the time it has, and tells you the arithmetic. If it does, raise the interval, split the building across two workflows, or send to fewer outlets.
- A restart interrupts the rollout. If Neowit restarts part way through, the remaining outlets aren't picked up again — they get their turn the next night.
- Anything else on that outlet reboots too. One outlet often feeds more than the screen.
- Don't stack this with a reactive reboot. If you also power cycle on a device-offline ticket, add a Cooldown so a room can't be rebooted twice in an hour.
Related
- Gude in workflows — the outlet actions in full.
- Power cycle a device that has gone offline — the reactive version.