Closed krishnanbalaji closed 8 years ago
Hi Balaji, this a design problem, which will be fixed (but not in too proper way) soon. You can give a try to netopeer2-cli, which should work much more reliably.
Regards, Michal
Thanks, Michael.
I haven't checked the code of netopeer2-cli. May be, I get an idea to change my client code.
Will get back.
Regards, Balaji On May 19, 2016 8:33 PM, "michalvasko" notifications@github.com wrote:
Hi Balaji, this a design problem, which will be fixed (but not in too proper way) soon. You can give a try to netopeer2-cli https://github.com/cesnet/netopeer2, which should work much more reliably.
Regards, Michal
— You are receiving this because you authored the thread. Reply to this email directly or view it on GitHub https://github.com/CESNET/libnetconf/issues/193#issuecomment-220352419
Hello, I am using only libnetconf just as a netconf client. My client has subscribed - ncntf_dispatch_receive() for notification. I have this situation where netconf server has crashed, but my notification receiver thread fails to get out of the loop in ncntf_dispatch_stop() which waits for someone set session->ntf_active to 0. This is a deadlock scenario. Seems to be a bug. Has anyone faced it?
Would appreciate a graceful way out rather than killing the respective thread through other means.
My relevant stack trace is like:
0 0x00007f9841ab0f2d in nanosleep () at ../sysdeps/unix/syscall-template.S:81
1 0x00007f9841ad9204 in usleep (useconds=) at ../sysdeps/unix/sysv/linux/usleep.c:32
2 0x00007f98414382f2 in ncntf_dispatch_stop (session=0x21d1020)
3 0x00007f984142b939 in nc_session_close (session=0x21d1020,
4 0x00007f984142cc01 in nc_session_receive (session=0x21d1020, timeout=0,
5 0x00007f984142d4a5 in nc_session_recv_msg (session=0x21d1020, timeout=0,
6 0x00007f984142d986 in nc_session_recv_notif (session=0x21d1020, timeout=0,
Regards Balaji