Egress
udServer records the outbound bytes of every response it serves, tagged with a category and — where the request belonged to one — an organisation, project and user. The Egress page in the Admin console reports that log server-wide. It requires the Analytics global permission.
Egress is what bandwidth actually costs, so this is the page that answers "why is the bill what it is". Organisations can see their own slice of the same data on their Usage tab; see the user guide's Organisations → Usage.
Categories
Every response falls into one of:
| Category | What it is |
|---|---|
| General API | Ordinary API traffic — listings, metadata, the web interface itself. |
| Stream data | Point cloud data streamed into udStream and other udSDK applications. |
| Downloaded data | Files downloaded to local storage. |
| Other | Anything not attributed to the above. |
The split matters because streaming and downloading are charged differently to organisations — see the user guide's Files.
Reading the Page
Choose a window of 7, 30 or 90 days. Within it you get:
- Total egress and the request count behind it.
- A breakdown by category, and a per-day chart so a spike can be placed in time.
- Top organisations by egress — the twenty highest, with an unattributed bucket for traffic outside any organisation (login pages, the documentation, and anything else that belongs to no organisation).
- Top users by egress — the twenty highest.
Two things to be aware of when reading the numbers:
- Names and email addresses in the user breakdown require the User List global permission. Volume is analytics; identifying who caused it is personal data, so the two are gated separately. Without User List the rows still appear, with the identifying columns withheld.
- An organisation or user that has since been deleted shows with an empty name. The volume is still counted — history is not rewritten when an account goes away.
Byte totals over a long window can exceed the range JavaScript integers represent exactly, so treat them as display figures rather than precise counts.
What is Not Counted Here
- Redirected reads. When a project has direct storage access enabled, the client fetches from S3 or Azure and those bytes never pass through udServer, so they cannot be recorded here. Your cloud provider's own billing is the record for those.
- Inbound traffic. Uploads are not egress and are not logged.
For the underlying API, see /_analytics/egress in the developer guide.