Search the Community
Showing results for tags 'docker'.
Found 4 results
So what is the "Best Practice" way to alert on memory usage inside of a docker container? I want to wrap an alert around something like cpu_percent_total but am not seeing a similar beast for memory. So next I am tempted to create a complex data point myself, but I found this post: that gives a formula to create my datapoint but also implies that it is of little value and that I should be alerting on swap usage to see issues with memory, but there is no swap datapoint in the Docker_Containers_cAdvisor20 DataSource. So I toss this debate to the crowd... Oh, Oracle of LogicMonitor, I beg your blessings in the form of sound advice on monitoring memory usage inside of a containerized environment that does not use any orchestration tools, and go...
We have multiple servers running Docker. Docker creates a network interface with the ip 172.17.0.0/16 by default, using 172.17.0.1 for the host. When we run a netscan policy, it finds the first server, and any subsequent servers running docker will be excluded as it detects that they all have the ip of 172.17.0.1 and assumes they are the same server. I want to be able to ignore some ips / ranges so the NSP does not incorrectly assume 2 devices are the same.
Hey All, Wondering if anyone here has had any experience monitoring hosts with docker containers? We're just starting to monitor some of our CoreOS hosts. SNMP is executing within a docker container on the host, and is mapping the physical /dev/sda disk to several file system mount points. For example (/etc/resolv.conf, /etc/hosts). Unfortunately this is not accurate, and I'm not sure why it mounts or detects those files as disks. It does accurately represent how much disk is being used, but I don't really need 5 'disks' reporting back the same data. Anyone have some ideas or thoughts on how I can get this cleaned up for just our CoreOS docker environment? Thanks!