We see an error every now and then while doing a REST call: "connection reset". See stacktrace below. What we know: - it happens with different endpoints, but only with one specific third party application, not (yet?) with others - most of the time calls to the same endpoints work fine, the error occurred 10 times in 25500+ calls over the last month - it only happens with POST, not with GET (at least we didn't see that yet) - there aren't any outgoing calls at the same time - the REST call does not return an httpResponse and no response code (like 500), only a latest error - the error happens immediately, within a few milliseconds (less than 10) - the log doesn't show anything else that may be connected with this, at least not on info/error level Do you have any pointers or possible solutions? Edit: There is this in the documentation, but I'm not sure if it applies, since we get the error immediately and not after some timeout Error calling REST service at Module.Microflow (CallRest : 'Call REST (POST)') [...] Advanced stacktrace: at com.mendix.integration.actions.microflow.RestCallAction.execute(RestCallAction.scala:79) Caused by: java.security.PrivilegedActionException: java.net.SocketException: Connection reset at com.mendix.integration.actions.microflow.RestCallAction.execute(RestCallAction.scala:79) Caused by: null at com.mendix.integration.actions.microflow.RestCallAction.execute(RestCallAction.scala:79) Caused by: Connection reset at com.mendix.integration.actions.microflow.RestCallAction.execute(RestCallAction.scala:79)
asked
Martin Koelewijn
1 answers
0
Hi Martin,
That sounds inconvenient indeed. I'm afraid there is no real solution here why the connection was broken. Theoretically, any switch in the network passed towards the endpoint and back could cause it. So the work-around shown in the documentation is probably the best option. Maybe you then can also log additional info when it occurs. Good luck!