Wintouch
Retour aux articles
Perspectives
Publié
Temps de lecture
9 min de lecture

Digital Signage Offline Playback Disaster Recovery: SLA

Digital signage offline playback disaster recovery: validate outage behavior, watchdog alerts, spare-player ratios and SLA tests before you sign.

Par Équipe d'ingénierie Wintouch
Digital Signage Offline Playback Disaster Recovery: SLA
Sur cette page

When the cloud CMS goes dark, the screen does not usually go black — it keeps looping whatever was last cached on the player, for as long as the local storage and the manifest allow. That default behavior is the whole purchase decision: whether the last frame is approved content, a stale price, an error page, or nothing at all. Vendors rarely volunteer the difference between those four outcomes, because "offline playback" is marketed as a checkbox rather than a failure mode with an owner, a latency and a written commitment.

This checklist converts offline playback into contract-verifiable terms across three distinct failure domains: your network drops, the vendor's cloud or your account becomes unavailable, and the hardware supplier can no longer ship replacements. Each item below ends in something you can test on site or demand in writing before signature.

What stays on screen: the three failure domains and their default behavior

Failure domain Default behavior if nothing is specified What you are actually buying
Store loses internet Player keeps playing the last synced playlist until cache or license expires; some players blank or show a splash screen A local manifest that does not need a live connection, plus a stated cache depth in days of scheduled content
Cloud CMS or account unavailable Content keeps playing, but scheduling changes, proof-of-play and remote commands stop; account suspension can disable a device entirely Written confirmation that playback is local-first, and that account or billing state cannot brick a player
Supplier cannot ship hardware Failed players stay dark until an RMA completes; unmatched spares sit in a warehouse with the wrong firmware Spare-player ratio by site tier, RMA clock in business days, and a field-replaceable unit (FRU) list

Signstream's operations guidance states the requirement plainly: reliable offline playback depends on preloading all assets and manifests, persistent local storage, and provisioning that survives a reboot — "get those three right, and an outage becomes invisible." That is the standard to hold every shortlisted vendor to, not the existence of an offline mode. See Signstream's offline playback essentials.

Checklist 1 — Local manifest, cache depth and switchover threshold

Ask for these five items in writing, then test them. None of the answers are universal; cache days, switchover latency and retention windows vary by player model and by how much video versus static content you schedule.

  1. Manifest independence. Does the player hold a local playback manifest that resolves without a network call, or does it re-fetch a manifest at each loop boundary? A player that resolves remotely will stall on the first loop after connectivity drops.
  2. Cache depth expressed in days, at your content weight. "64 GB storage" is not an answer. Ask: with our playlists and our asset sizes, how many days of scheduled content are resident locally, and what happens at day N+1 — freeze on last item, loop the oldest item, or show a fallback asset?
  3. Switchover threshold. How many failed heartbeats or seconds of unreachable CMS trigger local-only mode, and is that behavior configurable per site? Tie this to your alerting (Checklist 2) so the switch and the ticket happen on the same clock.
  4. Resync and reconciliation after recovery. When the link returns, does the player pull the current manifest and report what it actually played while offline? Without proof-of-play reconciliation, your campaign reporting under-counts the outage window and your schedule may re-run content that already aired.
  5. Time source. Players drift. Confirm NTP behavior when the store network is down and whether a real-time clock holds schedule logic through a power cut. A player that resumes at the wrong local time will fire the wrong daypart — a silent failure that looks like a content error.

Checklist 2 — Heartbeat alerting and watchdog behavior

Offline playback without monitoring just moves the failure somewhere quieter. DigitalSignage.com's maintenance and troubleshooting guide publishes a useful reference model: an alert table that treats a player offline for 5 minutes as a warning and 15 minutes as critical, alongside thresholds for CPU, memory, storage, temperature and playback error rates, with weekly log review for early warning signs. Those are that guide's recommended defaults, not an industry standard — use them as a calibration question: what are your defaults, who receives the alert, and how fast? Source: DigitalSignage.com troubleshooting guide.

  • Alert latency in writing. Time from last heartbeat to notification, and to which channel (email, SMS, ticket, webhook into your ITSM). Ask for the escalation path when the first contact does not acknowledge.
  • Silent-failure detection. Heartbeat covers a dead player. Ask how the vendor detects a player that is online but stuck — black output, frozen process, failed render — and whether playback-error counters feed the same alert stream.
  • Watchdog scope. Does a watchdog restart the player application only, or the whole OS? Does it survive a hard power cut and auto-resume playback without human login? Ask for the observed behavior, then verify it in the acceptance test.
  • Remote-control boundaries. Who can reboot, push content or wipe a device, and is every remote action logged? Operations teams need reboot rights at 2 a.m.; they do not need an unsupervised remote wipe path.

Checklist 3 — Cloud, account and license failures

This is the domain buyers forget to test, because demos never include a suspended account. Ask what happens on screen when the CMS subscription lapses, when an administrator account is disabled, when the vendor's status page shows a regional outage, and when an API credential expires. Envoy's cloud signage guide frames the right question for vendors: ask what continues to work if the service, your network, or a player goes down (Envoy).

Get three confirmations in the contract: playback is local-first and does not depend on an active license check at boot; account or billing state cannot disable devices in the field without a written notice period; and a documented breach or outage notification path exists with a named contact. If a vendor will not put the first item in writing, that single omission is usually enough to disqualify the platform for multi-site retail.

Checklist 4 — Spare players, RMA clocks and FRU boundaries

Hardware supply failure is where signage SLAs quietly collapse. A next-business-day RMA sounds adequate until you have forty stores and one depot. Ask instead for a ratio and a clock:

  • Spare-player ratio by site tier. Request the vendor's recommended ratio for flagship, standard and low-traffic sites, stated separately, and ask who holds the spares — you, the integrator, or the vendor. Ratios are not universal; they depend on fleet size, site criticality and travel time to site.
  • RMA turnaround in business days, measured from ticket acceptance, not from part dispatch, with the credit or penalty term attached if it is missed.
  • FRU boundaries. Which components can a store employee swap — player, power supply, cable, display — and which require a technician? Published FRU lists shorten outages far more than a faster depot. Where a display itself is the replaceable unit, insist on consistent mounting and touch interfaces across the fleet so a spare works in any store; our 23.8" industrial touch monitor and 32" industrial touch monitor are commonly specified as standard spares for this reason.
  • Firmware alignment. A spare with a different player image is not a spare. Ask how new units are staged to the fleet baseline before shipping.

Checklist 5 — Supply continuity and exit rights

Panel and board supply is a separate risk from player reliability, and it deserves its own clause set. Our display panel lead time and dual sourcing checklist walks through the questions to put to any hardware supplier. Three belong in the signage SLA as well:

  • End-of-life notice period for player and display models, long enough to phase replacements across stores.
  • Content and schedule export in a documented, non-proprietary format — playlists, schedules, asset files and proof-of-play logs — plus the contractual right to receive it on exit. Vendor lock-in is a disaster-recovery issue, not a procurement footnote.
  • Firmware and CMS migration path if the vendor's platform changes or the vendor is acquired.

The acceptance test script: three tests before you sign

Run these on one store, with the vendor's engineer present, using your own playlist and your own monitoring. Record timestamps; make the written report an exhibit to the contract.

  1. Pull the network cable for 30 minutes. Observe: does playback continue without a visible stutter, which asset plays, does the "offline" state reach local storage or trigger fallback content, and when does the alert arrive relative to the disconnection timestamp?
  2. Power-cycle the player mid-playlist. Observe: time to first frame after boot, whether login is required, whether content resumes at the correct schedule slot, and whether the CMS records the reboot.
  3. Disable the CMS account (or block the vendor endpoint) while the player stays online. Observe: does content keep playing, and does the player recover cleanly when access returns, including a manifest resync rather than a stale loop?

Add a fourth optional test if dayparting matters: change the player's clock or block NTP and confirm the schedule falls back safely. Any vendor who declines a witnessed run, or offers a demo environment instead of your site conditions, has told you something useful about the SLA negotiation ahead.

SLA clauses to get in writing

  • Offline fallback behavior, stated per failure domain, with the fallback asset named.
  • Alert latency and escalation path, with the monitoring channel specified.
  • Remote response window: time to first human response and time to remote remediation attempt.
  • RMA turnaround in business days and the remedy if missed.
  • Spare-player ratio by site tier and who holds the stock.
  • End-of-life notice period and content export rights on exit.
  • A named acceptance test report attached as an exhibit.

Reject any uptime percentage that arrives without a measurement method, monitoring source and exclusion list — an "online" dashboard that counts a looping black screen as healthy is not an uptime metric. Where a vendor cannot supply test evidence, the fallback is not a friendlier number; it is a written statement of behavior plus the acceptance run above.

Questions buyers ask at this stage

Does offline playback mean content updates stop? Yes — a change made while a store is offline appears only after reconnection. Pair the player with an operational backup for urgent location messaging, such as staff instructions or printed notices.

How many days of content should the cache hold? There is no universal figure. Ask the vendor to state it for your content weight and your longest realistic outage, and confirm what plays after that window closes.

Is a cloud-managed platform automatically a single point of failure? No, but only if playback is local-first and account state cannot disable devices. Get both in writing.

What spare ratio should we budget? Vendors and integrators quote different ratios for different site tiers; require their recommendation per tier, then test the RMA clock that backs it.

Request an offline-playback and spare-parts SLA review for your site list

Send us your store count, criticality tiers and the player or display model you are standardizing on, and we will review the failure-domain behavior, spare-parts plan and acceptance test items against your site list — including where a standard industrial display or touch monitor belongs in the spare pool, such as the 10.1" industrial touch monitor for counter and kiosk positions. Our digital signage solutions and media player range covers the hardware side of that review; the SLA conversation starts with the evidence you can test.

Request an offline-playback and spare-parts SLA review for your site list — bring the vendor's draft SLA and we will mark the clauses that still need a written commitment.

À propos de l'auteur

Équipe d'ingénierie Wintouch

Our commercial-display engineers and product team review specifications against current factory records, deployment requirements and published standards. Learn more about our capacité d’ingénierie et activité de fabrication.

Poursuivre vos recherches

Product facts and engineering claims follow our data and editorial policy.

Une question sur votre déploiement ?

Partagez votre environnement, la taille d'écran et les exigences d'intégration avec notre équipe d'ingénierie.