This message appears occasionally in our connector logs - anyone know what it means?
The timing of the messages roughly matches user complaints of connection instability but not definitively so.
[WARNING] [libsdwan] [client(999999…)] handle_message: incompatible user_flow_info objects:
Hi @dave54, what sort of connection instability are your users experiencing and how many are affected?
The line you reference should not by and of itself be a problem I think, but usually we need a complete set of logs and a problem description to help understand what might be going on.
@Emrul for context, I see this whenever I try to connect to my one hosted resource. It is a simple configuration, with access control set to the Everyone group.
What additional info can I provide in order to figure out what is going on?
@jkossis Ultimately I came to feel this line was not related to the troubles I was having.
The real culprit in my case was strict DNS timeouts while looking up *.twingate.com addresses. Once I switched my local DNS resolver to cache external addresses and use other available performance optimizations, my troubles went away.
It sounds like my application was similar to yours - I had just one hosted resource, and initially my local DNS configuration was to speak for only that one resource while forwarding all other requests to upstream resolvers. It turns out Twingate doesn’t really support that configuration.
It would have saved me and my users a lot of time and heartache if 1) they had just said so up front in the documentation; and 2) if the specific problem ("DNS Timeout") was more explicitly presented in the GUI vs. the misleading and all but useless "Connector offline" notifications / status (when it never was).