Monitoring Benefit
This is the OneDrive row of Service Level Availability. It reports what proportion of the selected period OneDrive was available, as measured by a file operation ENow performs against your tenant.
OneDrive is worth tracking separately from SharePoint even though they share a platform, because the impact differs. A SharePoint outage stops people reaching shared content. A OneDrive outage stops people reaching their own files — increasingly including desktop folders they do not think of as being in the cloud at all.
For how the figure is calculated, the two surfaces it appears on, the Node Type filter and where the thresholds live, see Configuration - Service Level Availability.
What feeds this figure
OneDrive availability is measured by a single round-trip test, which appears in the report's detail grid as OneDriveUploadDownloadDocument. It uploads a document and downloads it again.
That is a stronger test than a reachability check. It proves the service accepted a write, stored it, and returned it — which is what a user actually needs OneDrive to do. The trade-off is that it has more ways to fail, so read the error detail rather than assuming an outage.
Why users may not report it
This row has an unusual property: a short OneDrive outage often generates no complaints at all. The sync client holds files locally and retries quietly, so users keep working from the local copy and notice nothing.
That is a reason to take the figure seriously rather than dismiss it. Silence is not evidence the outage did not happen — it means the client absorbed it. The cost appears later, as sync conflicts and backlogs once connectivity returns.
The reverse is just as useful: a user reporting that "OneDrive is broken" while this figure is healthy is almost always describing a client problem — sync stopped, a conflict, a full disk, an account signed out — not a service one.
| What you see | What it means |
|---|---|
| Uptime dipped, nobody reported anything | Expected. The sync client masks short outages. The outage was still real. |
| Uptime healthy, a user says OneDrive is broken | Client-side. Look at that machine's sync status rather than the service. |
| OneDrive and SharePoint dip together | Shared platform. One incident, not two. |
| OneDrive dips alone | Narrower than a platform problem — check the test account's own OneDrive provisioning and licence. |
Reading the failure detail
Because the test both writes and reads, an Elapsed Milliseconds of 0 is meaningful: the operation never started, which usually points at authentication or a token failure rather than storage. A high elapsed time before a failure points the other way — the transfer began and did not finish, which is a network or throttling story.
Common warning or error results, and potential solutions
| Result | Potential solution |
|---|---|
| Sustained 0% uptime | Usually collection or the test account. Confirm the account is licensed and that its OneDrive has actually been provisioned — an unprovisioned OneDrive fails permanently rather than intermittently. |
| Failures with zero elapsed time | The operation never began. Check authentication and the app registration's permissions before looking at OneDrive itself. |
| Failures confined to one Site | That location's path, not the service. Filter by Node Type to confirm. |
| Uptime drops after a licensing change | Check the monitoring account still holds a licence that includes OneDrive. |
| Uptime recovers but users report missing files | Sync conflicts from the outage window, not a continuing service problem. Work the affected clients. |
Comments
0 comments
Article is closed for comments.