Please, Dojo,
When a protocol makes use of 2 (or more) concurrent interfaces, e.g. SNMP 161 & http 8080, and one is blocked for any reason (e.g. fw policy to be updated or target configured to use a different http port), what's the default behaviour for a timeout event?
Would I always get a "global" element timeout when either of the two ports is not receiving data?
Or is the other interface somehow able to act as a "keep-alive" for the other one at "element" level?
In other words, when looking at the streamviewer, if the SNMP agent is no longer working on the target device, would I get a timeout even if my http polling is still successful?
Hi Alberto,
There is a small note in DataMiner Docs related to this topic:
- Prior to DataMiner 9.5.6, if an element has multiple connections, it will not go into timeout state as long as one of the connections is still communicating properly. From DataMiner 9.5.6 (RN 16909) onwards, this is no longer the case: as soon as one of the connections times out, the element will go into timeout.
Users can clear the Include timeout option for a particular connection while creating or editing an element to ensure that this connection does not influence the overall timeout state of the element.
Top notch Miguel – I didn’t know this change had happened.
I was a bit of an old-school timeout-er 🙂
I’ll play around with the option – and possibly will come back with a few more questions – your feedback is always great! Thank you