I just upgraded two heavy forwarders from DB Connect 3.0.0 to 3.1.3. On both forwarders, I can create a new input and save it. But if I try to clone an input and save, or edit an existing input and save, the "Finish", the screen hangs for 30-60 seconds before an orange banner shows the error "Failed to fetch". The input never saves. Does not appear to be driver-specific.
I am able to use the CLI to edit db_inputs.conf.
I don't see any errors in the internal logs, not sure if boosting the logging level will show an error? Has anyone seen this?
To be fair, I had a similar problem saving before, while using DB Connect 3.0.0, but only about half of the time. In 3.0, there was no error visible, except when I click back to the query screen I would see good query results, but an "undefined" error was displayed below the results...
Adding an answer, although this post is old...
This appeared to be an issue between Splunk and Chrome. Using Chrome debugger, I could see that Splunk may have been attempting to open more than 6 connections (sockets??), and Chrome only allows 6. This error did not occur in IE or Firefox, which both allow more connections, like ~13.
We updated Splunk to the latest version at the time, which I think was 7.1. The problem persisted. I made sure to use Firefox after this, when making edits in DB Connect. Eventually, I tried again in Chrome and there is no issue. Not sure if an update to Chrome fixed this, or an update to DB Connect, or something else.
Adding an answer, although this post is old...
This appeared to be an issue between Splunk and Chrome. Using Chrome debugger, I could see that Splunk may have been attempting to open more than 6 connections (sockets??), and Chrome only allows 6. This error did not occur in IE or Firefox, which both allow more connections, like ~13.
We updated Splunk to the latest version at the time, which I think was 7.1. The problem persisted. I made sure to use Firefox after this, when making edits in DB Connect. Eventually, I tried again in Chrome and there is no issue. Not sure if an update to Chrome fixed this, or an update to DB Connect, or something else.