Open wudyw opened 7 months ago
I think this issue is caused by timeout in nc_recv_reply() is too small, I modified it to 10ms, it is much easier to trigger, then I modified it to 1s, didn't see any error. I wonder if timeout, why the ret is not NC_MSG_WOULDBLOCK /waiting for another message timed out /, is it not timeout?
What libnetconf2 version are you using? What NETCONF client and server are you using? I do not understand that from waht you wrote. And what nc_recv_reply()
timeout are you talking about, can you mention the project/source file?
What libnetconf2 version are you using? What NETCONF client and server are you using? I do not understand that from waht you wrote. And what
nc_recv_reply()
timeout are you talking about, can you mention the project/source file?
libnetconf2 I used is 3.0.8, I just used libnetconf2 as a client, the server is Nokia router.
r = nc_recv_reply(session, rpc, msgid, 100, &envp, &op);
The time out is 100, the 4th para in nc_recv_reply
What libssh version are you using? Based on what you said, it seems ssh_channel_poll_timeout()
is returning something it is not supposed to and there was a bug a long time ago that SSH_AGAIN
was returned even though it should not have been.
I created a session using libssh, but sometimes it print this log, I SET the timeout is 100ms, the timeout is too small to lead this issue?
//log