In the PVA log, there are resource usage alert messages about the
T=10:25:24:056; L=(inf); PID=57797; TID=7f3cba2e7760; P=VZLAlertManagerLocal [timeoutEvent] Send resource alert type 2 eid eb5ce675-7362-0545-9968-d327190a5ff2 counter counter_cpu_share_used (cur=100, hard=95, soft=85)
What does this message mean and how do I get rid of it?
This message means that either the
red_alerts_threshold has been reached for the container CPU limit - there was a load spike inside the container.
counter_cpu_share_used - the real CPU usage of a container against the CPU limit set for it.
In order to get rid of these messages, either change the alert threshold values or increase container CPU limit:
red_alerts_thresholdvalues as per this article:
To change container CPU limit, use PVA Control Center, Parallels Management Console, or command line tools:
vzctl set CTID --save --cpulimit 200
For more information regarding CPU limits in Parallels Virtuozzo Containers, refer to the following articles:
Adjusting CPULIMIT parameter for Parallels Virtuozzo Containers
CPU Guarantee and CPULIMIT configuration
CPU limits design in Parallels Server Bare Metal 5, Parallels Cloud Server and Parallels Virtuozzo Containers for Linux
For information about
counter_memory_share_used alert see:
counter_memory_share_used resource alert mean in PVA?