As per code, it is difficult to predict why the brick was crashed in this
function but I think we can avoid this crash after saving the
host_name/base_path
in the starting of this function(posix_fs_health_check). Though we do
cancel this health check thread at the time of calling posix_fini and cal
I am also seeing a bunch of these errors in the logs. Do they really need
to be Info logs?
[2018-11-14 06:01:18.597029] I [socket.c:693:__socket_rwv]
0-tcp.patchy-vol01-server: EOF on socket 382 (errno:61:No data available);
returning ENODATA
[2018-11-14 06:01:18.597640] I [socket.c:693:__socket
On 11/14/2018 10:04 AM, Nithya Balachandran wrote:
> Hi Mohit,
>
> The regression run in the subject has failed because a brick has crashed inĀ
>
> bug-1432542-mpx-restart-crash.t
>
>
> *06:03:38* 1 test(s) generated core
> *06:03:38* ./tests/bugs/core/bug-1432542-mpx-restart-crash.t
> *06:03:
Hi Mohit,
The regression run in the subject has failed because a brick has crashed in
bug-1432542-mpx-restart-crash.t
*06:03:38* 1 test(s) generated core *06:03:38*
./tests/bugs/core/bug-1432542-mpx-restart-crash.t*06:03:38*
The brick process has crashed in posix_fs_health_check as this->pri