[ClusterLabs] corosync eats the whole CPU core in epoll_wait() on one node in cluster

[Andrew Beekhof]

> On 2 Jun 2015, at 12:20 am, Jan Friesse <jfriesse at redhat.com> wrote:
> 
> Andrew Beekhof napsal(a):
>> 
>>> On 1 Jun 2015, at 5:21 pm, Jan Friesse <jfriesse at redhat.com> wrote:
>>> 
>>> Vladislav
>>> 
>>> Vladislav Bogdanov napsal(a):
>>>> 
>>>> nearest socket inode (connecting one, 17986) is used by pacemakerd.
>>>> 
>>>> strace of pacemakerd shows absolutely normal
>>>> poll([{fd=8, events=POLLIN}, {fd=6, events=POLLIN}, {fd=5,
>>>> events=POLLIN}, {fd=4, events=POLLIN|POLLPRI}], 4, 500) = 0 (Timeout)
>>>> poll([{fd=8, events=POLLIN}, {fd=6, events=POLLIN}, {fd=5,
>>>> events=POLLIN}, {fd=4, events=POLLIN|POLLPRI}], 4, 500) = 0 (Timeout)
>>>> poll([{fd=8, events=POLLIN}, {fd=6, events=POLLIN}, {fd=5,
>>>> events=POLLIN}, {fd=4, events=POLLIN|POLLPRI}], 4, 500) = 0 (Timeout)
>>>> 
>>> 
>>> AFAIK pacemaker doesn't use libqb qb_loop but corosync does. That's the reason.
>> 
>> The reason for?
> 
> strace of pacemakerd shows absolutely normal.

ah, now i see that line lurking above the strace output