Subject: Unable to reconnect after forcible disconnect - Most likely a DotNetty issue.
Hi,
First let me say thank you for such a complete implementation of XMPP in .NET. I've been playing with using MatriX.vNext to connect and "listen" to events on a Cisco Finesse server. I was hoping to wrap my implementation inside of a "set it and forget it" Windows service. Upon testing it's ability to respond to "adverse network conditions" (I unplugged my Ethernet cable which triggered the connected socket to be "forcibly closed") I found that attempting to call ConnectAsync a second time results in a never-ending wait for the socket to reconnect. Even plugging the network back in does not allow the application to continue. It is essentially "dead in the water."
I did go to the GitHub project for DotNetty and look into this issue and found the following issue already opened by another:
https://github.com/Azure/DotNetty/issues/245
I also added a comment to the end of the chain explaining that the recommendation of adding a ChannelHandler to catch the ChannelInactive or ExceptionCaught events and attempting to reconnect within does not work either. The same never-ending wait appears to occur which is what leads me to believe this is a DotNetty issue and not a MatriX.vNext issue.
Has anyone else experienced this? This is a real show stopper for me in that I cannot write a reliable service that will simply re-connect upon a temporary network issue and the service will require a manual (most likely system administrator initiated) restart to correct.
First let me say thank you for such a complete implementation of XMPP in .NET. I've been playing with using MatriX.vNext to connect and "listen" to events on a Cisco Finesse server. I was hoping to wrap my implementation inside of a "set it and forget it" Windows service. Upon testing it's ability to respond to "adverse network conditions" (I unplugged my Ethernet cable which triggered the connected socket to be "forcibly closed") I found that attempting to call ConnectAsync a second time results in a never-ending wait for the socket to reconnect. Even plugging the network back in does not allow the application to continue. It is essentially "dead in the water."
I did go to the GitHub project for DotNetty and look into this issue and found the following issue already opened by another:
https://github.com/Azure/DotNetty/issues/245
I also added a comment to the end of the chain explaining that the recommendation of adding a ChannelHandler to catch the ChannelInactive or ExceptionCaught events and attempting to reconnect within does not work either. The same never-ending wait appears to occur which is what leads me to believe this is a DotNetty issue and not a MatriX.vNext issue.
Has anyone else experienced this? This is a real show stopper for me in that I cannot write a reliable service that will simply re-connect upon a temporary network issue and the service will require a manual (most likely system administrator initiated) restart to correct.