I've been testing on my branch that is kept pretty close to grpc/grpc master.ĭoes anyone have any idea what is going on? I have the bad process open on a test server in gdb and have timeouts implemented so I can run repeated requests at it to continue debugging if necessary.
#Yep message core full#
I have the full logs if they might be helpful. I've been working on this for a few weeks now and haven't been able to track down why or a fix. (and a whole lot more)Įffectively it seems like the something is not getting scheduled when it should be (grpc_combiner_continue_exec_ctx never gets run). I've managed to get GPRC_TRACE=all GRPC_VERBOSITY= DEBUG of a good request and a request when it is in the bad state to compare side by side. The stack trace doesn't always look like that but it always is stuck on epoll_wait/epoll_pwait often from gpr_cv_wait. Tag=0x0) at src/core/lib/surface/server.c:1486
0 kommentar(er)
