1

I've been seeing instances of 'fatal_network_error' popping up in production more frequently recently.

This particular error unfortunately isn't documented anywhere as to what the issue may be, so apart from retrying the doc in the client open I'm unsure of how to prevent something like this from consistently happening to people. In particular, one document [ID: 0B1es-bMybSeSb3hnSHdRUTNLSGc] that I have access to appears to be hitting this nearly 100% of the times I have opened it in Firefox (though it appears this is not exclusive to FF from other error reports).

After re-requesting the same document to be opened a second time (after handling the fatal error), the document load is successful. Looking at the network requests identical requests are made with the exact same params (rctype,recver,id,access_token). Additionally the response from the server is identical apart from what looks like a version hash and possibly a version number after the Document ID.

Is this a known issue? Are there workarounds?

I have shared the document publicly, please let me know if there is any other information that I can provide that may help track this down.

Stack in drive DEBUG build.

Grant Watters
  • 660
  • 4
  • 16
  • Can you show your code? It looks like variable `a` is not define in the initial call. But if you check the [document](https://developers.google.com/google-apps/realtime/reference/gapi.drive.realtime.Error), fatal errors cannot be recovered from and require the document to be reloaded. You have to create a function that will catch this error and reload the document. Hope this helps! – Mr.Rebot Aug 27 '16 at 15:22
  • Thanks for your reply! The undefined variable is deep in the Realtime API library which unfortunately I do not have direct control over. I am handling fatal errors as described in their documentation which is how I am able to successfully reload the document on the second attempt. – Grant Watters Aug 27 '16 at 16:41

0 Answers0