Browse the documentation
Gallery Traffic: Tube Sources
Applies to: Publishers · Ad manager setups only, when the network has the module switched on
A tube source is the video dump your own tube produces: a CSV, JSON or XML file with one row per video. The panel copies it on a schedule and keeps its copy in sync, so your partners always get your current catalogue. You keep producing the dump; the panel only reads it.
Adding a source#
Open Publisher → Gallery Traffic → Tube sources and click Add source.
| Field | What to enter |
|---|---|
| Name | Your own label, for example the tube's name. |
| Tube domain | Your tube's domain, such as yourtube.com. Page links may only point to this domain or its subdomains. |
| Dump URL | The full address of the dump. It must be a public http:// or https:// address; private and internal addresses are refused. |
| Format | CSV, JSON or XML. Use JSON if your tube can produce it: named fields need no guessing. |
| Refresh every (minutes) | How often the copy is refreshed. At least 5. |
| Column separator / List separator | CSV only: the character between columns, and the one between items inside a column, such as categories. |
| Active | Untick to stop copying without deleting the source. |
Click Copy now to pull it straight away instead of waiting for the schedule.
What the panel reads from each row#
Each row needs at least a video id and a URL; rows without both are dropped, because there is nothing to track and nothing to send the visitor to. Common names are understood: id or videoid for the id, link for the URL, post_date for the date added, screenshot or poster for thumbnails, plus title, description, categories, duration and quality. A duration like 12:30 is read as minutes and seconds.
A CSV without a header row is fine. The panel checks which column really holds the links, ids, images and dates instead of assuming an order, and refuses the pull if it can't find an id and a URL, rather than storing a wrong catalogue.
Placeholders your tube puts in its own video URLs, such as {siteid}, {pubid}, {feedid} and {sourceid}, are filled in with the partner's values when the visitor is sent on. Placeholders it doesn't know are left as they are.
Pull status#
The source list shows Last copy, Videos and the result of the last pull:
| Status | Meaning |
|---|---|
| up to date | The dump was copied; new videos were added and removed ones were retired. |
| came back empty | The dump returned no videos. Nothing was removed, since a broken URL is far likelier than an empty tube. Check the URL. |
| copied, removals held — URL is capped | The dump returned exactly as many rows as its own limit asks for, so videos past the cap are missing but still live on your tube. Nothing was removed. Fix the URL: raise or drop the limit. |
| copied, removals held | More than a fifth of the catalogue (and at least 25 videos) would have been removed in one go. Nothing was removed. If you really took those videos down, click Sync removals. |
| The pull failed | The dump could not be fetched or read. The message says why. |
| never pulled | The source is new. Click Copy now. |
Removals are held this carefully because a removed video goes out in every partner's removed-videos list, which tells them to take that page down on their site.