Since Sunday evening, 11 October, I have no longer been able to refresh my Power BI report
via Bridge Online (odata4). Power BI gives:
TransactionLinesIncremental@eol: “OData: Invalid JSON. Unexpected end of input was found in JSON content.”
SalesInvoiceLinesIncremental@eol: “Exception from HRESULT: 0x80040E4E” (follow-up error)
In Session I/Os I see the following for division 1033176:
• The first page comes in from the cache within a few ms.
• The follow-up page hangs for 23 seconds each time and returns 0 bytes, always with the same skiptoken:
This is retried about 30 times per table, after which the Bridge Online request arrives at Power BI as aborted (after roughly 400 seconds).
Divisions 2338104 and 3161833 do work normally.
Timestamps (as shown in Session I/Os): 11-10-2026 20:19 – 20:53 and again 21:53 –
21:54.
At 21:50 I performed a “Cache Reset” on bridge-online.cloud, but afterwards
exactly the same thing happens: first page from cache, then the same
skiptokens again with 23 seconds and 0 bytes.
It looks as though the incremental storage for division 1033176 has got stuck at a fixed
position.
Could you reset this for this division, or show me how I can fix this myself?
I can also upload the export of the Session I/Os if needed.
The Session I/O shows that the first page comes from the cache and that the page with $skiptoken ends after 23 seconds without a response. This happens every time with the same $skiptoken. This fits with a sync API of Exact Online that does not respond to that request within the time limit. Resetting the cache in Bridge Online does not change this.
Bridge Online retries the request several times. Power BI aborts the download after approximately 400 seconds. This results in the messages “Unexpected end of input” and 0x80040E4E. These messages are consequences. They say nothing about the cause.
Invantive cannot resolve this on Invantive’s side. The following steps are possible:
Contact Exact Online Support and refer to the Session I/O. Provide the URL, the $skiptoken and the time. In similar reports, Exact Support has confirmed the cause (see below). It is possible that Exact will insist on a different subscription model.
Use TransactionLinesBulk instead of TransactionLinesIncremental, and SalesInvoiceLinesBulk instead of SalesInvoiceLinesIncremental. The Bulk tables support server-side filtering, for example by financial year or period. Whether they work without a timeout depends on Exact Online. This is not guaranteed.
The tables with Incremental always process all rows of the division, even with a filter applied. With a large division, this increases the likelihood of a timeout.
If the problem persists, the information from Monitoring can help. That is the status code and the duration of the request, and whether Exact Online or the connection closes the request.
Deze vraag is automatisch gesloten na 1 week inactiviteit. Het laatste gegeven antwoord is gemarkeerd als oplossing.
Gelieve een nieuwe vraag te stellen via een apart topic als het probleem opnieuw optreedt. Gelieve in de nieuwe vraag een link naar dit topic op te nemen door de URL er van in de tekst te plakken.