I 'm using corosync
and OS Debian. When I start qpidd program write this message:
Warning ClasterClock task late 1 times by 70ms on average
I 'm using corosync
and OS Debian. When I start qpidd program write this message:
Warning ClasterClock task late 1 times by 70ms on average
Your question is quite unclear, but to synchronize time on any herd of Debian machines, you set up the ntp
daemon on one server to be in server mode and point ntp
on the rest of the machines to it.
This actually doesn't have anything to do with time synchronization. Instead, that warning means that the broker was expecting to run the ClusterClock task at time t, but instead it tends to run, on average, at time t + 70ms. If the broker is busy doing other things (e.g. processing messages), you may see tasks such as this one firing later than planned. If desired, you can change the frequency of the ClusterClock task by setting the "cluster-clock-interval" option (the default is every 10ms).