Sync Status & Troubleshooting
Check connection status, and spot and fix sync failures
How to tell whether a data pool is syncing cleanly, and what to do when it isn't. Setup and configuration of a data source: Data Sources & RAG.
Connection status
Each data source in the data source list shows a status. The line below it names the exact reason.
| Status | Meaning |
|---|---|
| Up to date | All files are read in and searchable. |
| Searchable | All files can be found by keyword; the meaning index is still being built. |
| Indexing | A sync is running, waiting in the queue, or being retried. |
| Needs attention | Something needs a person, e.g. an expired sign-in, an unreachable source, or a stalled sync. |
| Paused | The data source is paused. |
| Unknown | The status cannot be determined right now, e.g. because an Outpost has not sent a current report. |
You see the connection of an Outpost itself on its card under Settings → Outposts. If it is reachable but not fully working, the card shows Reachable · limited operation with the reason below. What each display means: Outpost connection status.
Common causes when a single data source shows Needs attention:
- The source's access token (e.g. SharePoint, Google Drive, Confluence) has expired or was revoked
- Permissions were revoked in the source system
- Network or firewall changes on the source system's side
- For an Outpost data source, the status reflects reachability from MeinGPT's side - it can differ from the Outpost's own local indexing state. If the Outpost console shows the folder fully indexed but MeinGPT still shows Needs attention or Unknown, check the Outpost's own connection display for the underlying cause.
Attention
If all of your data sources show "Needs attention" with the reason "The source could not be reached" at the same time — including sources with completely different underlying systems (e.g. SharePoint, Google Drive, and Confluence simultaneously) — this usually does not point to a problem with the individual sources, but to a temporary platform-side issue. In that case, contact MeinGPT support instead of reconnecting individual sources.
Monitor sync status and fix failures
Every data pool syncs automatically every 15 minutes. If a sync fails, you can see it and fix it directly in MeinGPT.
Automatic alert email
Admins automatically get an email when there's a problem with a data pool:
- Sync failing – a data pool has failed to sync several times in a row. New or changed documents stop being indexed until this is fixed.
- Data source connection lost – an OAuth connection (e.g. SharePoint) repeatedly fails to refresh its access token.
- Spike in sync errors – an unusually large share of files errored out during a sync.
Each email links directly to the affected data source in your settings. You can toggle these notifications individually under Settings → Notifications → Outpost alerts.
Check sync status
- Open Settings → Data sources
- Click the affected data source
- The Activity tab lists the latest syncs. Each row shows the time, the result (Running, Queued, Succeeded, Failed or Interrupted) and, if some files were not searchable, how many. Consecutive syncs without changes are folded into one row.
Open a failed or interrupted sync to see Cause, Consequence, What to do and Next attempt. Below that is a support code you can copy with Copy for support. Failed syncs are retried automatically, and the previous state stays searchable.
Which individual file is missing and why is shown in the Files tab (next section). How many files are searchable in total is shown on the Overview (Indexing progress). The Files and Activity tabs exist only for Cloud data sources; an Outpost data source shows the Overview only.
Individual files missing: reasons in the Files tab
The Files tab on the data source's detail page groups every file that is not searchable by reason and names the next step for each group. Under Check you see how far each file got: found, read, split into passages, keyword search, meaning search.
| Label | Meaning | What to do |
|---|---|---|
| Too large | The file exceeds a processing limit. This can be file size, but also the page count of a scanned PDF (more than 50 pages, see Limits per file) or an extreme amount of text. | For a scanned PDF: add a text layer with an OCR tool. Otherwise: split the file or make it smaller. It is added automatically at the next sync. |
| No text found | An empty file, or a scan without recognisable text. | Store a version with text, such as a searchable PDF. |
| Reading failed | An error occurred while reading that fits no other group. | Start a new sync. If the error stays, send the support code (see below) to support. |
| Timed out | Reading took too long. | Start a new sync. If it stays like this, send the support code. |
| No read access | The connected account may not read the file. | Give the account read access, or reconnect the source with an account that may read. |
| File damaged | The file cannot be opened, for example a truncated Word document. | Store an undamaged copy. |
| Password-protected | MeinGPT cannot read encrypted files. This includes password-protected files and files to which a sensitivity label (Microsoft Purview / MIP) applies encryption. | Remove the password protection or store an unprotected version. For an encrypting sensitivity label: put the content into a separate library or folder whose files carry a label without encryption, agreed with your IT. |
| Format not supported | MeinGPT cannot read this file as one of the supported formats (see Formats and access). | Nothing to do for the file itself. Store the content in a supported format, such as a PDF or Office file, if it should be found. |
| System file | Temporary and system files are skipped on purpose. | Nothing to do. |
Support code. For a file, MeinGPT shows a code of the form reason.detail · ksref_…, for example too_large.reader_limit · ksref_…. Copy for support copies it so you can give it to support. It points to the data source and the run, but contains no file content.
Progress and "Stuck"
While a data source is being indexed, the Overview shows three steps with a counter each:
| Step | What is counted |
|---|---|
| Reading files | How many of the found files have been read. |
| Keyword search | How many files can be found by keyword, out of all that can become searchable at all. |
| Semantic search | How many of those can also be found by meaning. |
Files that could not be read and skipped files count only at Reading files, not in the two search steps. The search counters can therefore sit below the total of the file row without anything missing: the Overview lists those files separately as not searchable. The Semantic search row only appears if a meaning index is built for the source. Outpost data sources report totals only, so they have neither the queue nor Stuck.
If a sync is waiting for a free slot, the page shows Waiting for a free slot · place N in the queue. That is not an error; the list shows the data source as Indexing.
Stuck appears when files or index work are still outstanding and nothing has moved for more than 24 hours: no file changed state, the meaning index finished nothing, and no sync started. A large first sync does not count as stuck while it keeps moving files. The notice reads Stuck: the last movement was …, and the list shows the data source as Needs attention.
What to do:
- Start a new sync with Sync now. If it moves files forward, the notice disappears.
- Open the Files tab and check whether the missing files are listed there with a reason you can fix yourself.
- If it stays that way, contact support@meingpt.com with the data source's name, the time after "last movement" and the three counters.
A sync fails completely
If the Activity tab shows a sync as Failed, Cause names the reason and What to do the next step. This is a different kind of failure than a single file that is not searchable: the sync itself failed, for example because the sign-in expired or the source is set up incorrectly or cannot be reached.
Typical causes for a Cloud data source such as SharePoint/OneDrive, Google Drive, or Confluence:
- the configured site, drive, or folder was moved, renamed, or deleted in the source
- the permission on that specific site or library was narrowed afterwards in the source system, even though the OAuth connection itself is still valid
- a temporary outage or rate limit on the source system's side
First check whether the connected site/library still exists in the source system under the same address, and whether the permission there still allows access to exactly that scope - not just general access to the source system. If triggering a new sync (below) doesn't help, contact support@meingpt.com with the data source's name, the time of the failed sync and the support code.
Manually trigger a resync
- Open the affected data pool under Settings → Data sources
- Click Sync now at the top to start a new sync immediately
Every sync run – whether automatic or manually triggered – also removes entries for files that have been deleted from the linked source. A manual resync is therefore the right way to clear out stale index entries without having to recreate the data source. In rare cases, a single deletion may be missed during the sync and will still be present after a resync – if that happens, reach out to MeinGPT support so we can look at the affected data source directly.
Re-authenticate the connection (OAuth sources)
- Open the affected data pool under Settings → Data sources
- Click Reconnect and authenticate again
- Your existing source settings stay unchanged — only the authorization is refreshed
If that doesn't help, contact support@meingpt.com.
"Could not verify account" during initial setup
This message can appear the very first time you connect an OAuth source, before any sync has run - usually because the Microsoft or Google sign-in window closed without returning a valid authorization code. For SharePoint, the most common cause is missing admin consent for the MeinGPT application in your Microsoft tenant; details and other causes: SharePoint/OneDrive troubleshooting.