Re: sshd-core: dealock

2016-12-05 Thread Markus Rathgeb
Hello again, I am still trying to create a minimal example to reproduce the problem. If I am using only 1 NIO worker on server side it is easy to reproduce. So, I assume there must be at least two workers, that there is one that always could call the notifyAll method to wake up all the sleeping

sshd-core: dealock

2016-12-02 Thread Markus Rathgeb
Hello, I have a project that is using sshd-core (+mina-core 2.0.16) on Client and Server side. I previously used 1.3.0 and switched to 1.4.0-SNAPSHOT. The problem (I will describe shortly) is raised by both versions (not pre 1.3.0 version has been used before). The clients logic is very simple: