ganglia / monitor-core

Ganglia Monitoring core
BSD 3-Clause "New" or "Revised" License
491 stars 246 forks source link

Ganglia Missing Metrics for Some Nodes #308

Open suanmiao opened 6 years ago

suanmiao commented 6 years ago

We are using Ganglia (gmond 3.6.0) and there are some metrics missing for some nodes. Explanation or suggestion would be appreciated regarding this issue. Thanks!

Below is the detail

Symptom

The metric load_one (which shows our average one minute CPU load) is partially or completely missing for some nodes.

image

And the metric reported from other nodes during the same period looks like this:

image

And we checked the rrd file under /var/lib/ganglia/rrds, it shows that some rrd files are missing for these nodes with issue.

Below are the files under that folder for a normal node:

boottime.rrd bytes_out.rrd cpu_idle.rrd cpu_num.rrd cpu_system.rrd cpu_wio.rrd disk_total.rrd load_five.rrd mem_buffers.rrd mem_free.rrd mem_total.rrd pkts_in.rrd proc_run.rrd swap_free.rrd bytes_in.rrd cpu_aidle.rrd cpu_nice.rrd cpu_speed.rrd cpu_user.rrd disk_free.rrd load_fifteen.rrd load_one.rrd mem_cached.rrd mem_shared.rrd part_max_used.rrd pkts_out.rrd proc_total.rrd swap_total.rrd

Below are the files under the node with issue:

boottime.rrd cpu_num.rrd cpu_speed.rrd mem_total.rrd swap_total.rrd

Thus we conclude that Ganglia failed to record these metrics.

Setup & Environment

Ganglia Version: gmond 3.6.0 System Version:

Gmond.conf:

/* Ganglia modules are defined in terms of .conf files in /etc/ganglia/conf.d; the directive below includes all such .conf files.

Ganglia Python modules (e.g. the NVIDIA python module for monitoring GPUs) are specified in .pyconf files within /etc/ganglia/conf.d. When we install support for Ganglia python modules, (via apt-get install ganglia-monitor-python in our GPU base image) the installation process creates a /etc/ganglia/conf.d/modpython.conf file. The directive below includes the modpython.conf file, which in turn contains a directive to include all .pyconf files within /etc/ganglia/conf.d. / include ('/etc/ganglia/conf.d/.conf')

/ This configuration is as close to 2.5.x default behavior as possible The values closely match ./gmond/metric.h definitions in 2.5.x / globals { daemonize = yes setuid = yes user = ganglia debug_level = 0 max_udp_msg_len = 1472 mute = no deaf = no host_dmax = 0 /secs / cleanup_threshold = 300 /secs / gexec = no send_metadata_interval = 0 }

/ The host section describes attributes of the host, like the location / host { location = "unspecified" }

/ You can specify as many tcp_accept_channels as you like to share an xml description of the state of the cluster / tcp_accept_channel { port = 8649 }

/ Each metrics module that is referenced by gmond must be specified and loaded. If the module has been statically linked with gmond, it does not require a load path. However all dynamically loadable modules must include a load path. / modules { module { name = "core_metrics" } module { name = "cpu_module" path = "/usr/lib/ganglia/modcpu.so" } module { name = "disk_module" path = "/usr/lib/ganglia/moddisk.so" } module { name = "load_module" path = "/usr/lib/ganglia/modload.so" } module { name = "mem_module" path = "/usr/lib/ganglia/modmem.so" } module { name = "net_module" path = "/usr/lib/ganglia/modnet.so" } module { name = "proc_module" path = "/usr/lib/ganglia/modproc.so" } module { name = "sys_module" path = "/usr/lib/ganglia/modsys.so" } }

/ The old internal 2.5.x metric array has been replaced by the following collection_group directives. What follows is the default behavior for collecting and sending metrics that is as close to 2.5.x behavior as possible. /

/ This collection group will cause a heartbeat (or beacon) to be sent every 20 seconds. In the heartbeat is the GMOND_STARTED data which expresses the age of the running gmond. / collection_group { collect_once = yes time_threshold = 20 metric { name = "heartbeat" } }

/ This collection group will send general info about this host every 1200 secs. This information doesn't change between reboots and is only collected once. / collection_group { collect_once = yes time_threshold = 1200 metric { name = "cpu_num" title = "CPU Count" } metric { name = "cpu_speed" title = "CPU Speed" } metric { name = "mem_total" title = "Memory Total" } / Should this be here? Swap can be added/removed between reboots. / metric { name = "swap_total" title = "Swap Space Total" } metric { name = "boottime" title = "Last Boot Time" } metric { name = "machine_type" title = "Machine Type" } metric { name = "os_name" title = "Operating System" } metric { name = "os_release" title = "Operating System Release" } metric { name = "location" title = "Location" } }

/ This collection group will send the status of gexecd for this host every 300 secs / / Unlike 2.5.x the default behavior is to report gexecd OFF. / collection_group { collect_once = yes time_threshold = 300 metric { name = "gexec" title = "Gexec Status" } }

/ This collection group will collect the CPU status info every 20 secs. The time threshold is set to 90 seconds. In honesty, this time_threshold could be set significantly higher to reduce unneccessary network chatter. / collection_group { collect_every = 20 time_threshold = 90 / CPU status / metric { name = "cpu_user" value_threshold = "1.0" title = "CPU User" } metric { name = "cpu_system" value_threshold = "1.0" title = "CPU System" } metric { name = "cpu_idle" value_threshold = "5.0" title = "CPU Idle" } metric { name = "cpu_nice" value_threshold = "1.0" title = "CPU Nice" } metric { name = "cpu_aidle" value_threshold = "5.0" title = "CPU aidle" } metric { name = "cpu_wio" value_threshold = "1.0" title = "CPU wio" } / The next two metrics are optional if you want more detail... ... since they are accounted for in cpu_system. metric { name = "cpu_intr" value_threshold = "1.0" title = "CPU intr" } metric { name = "cpu_sintr" value_threshold = "1.0" title = "CPU sintr" } / }

collection_group { collect_every = 20 time_threshold = 90 / Load Averages / metric { name = "load_one" value_threshold = "1.0" title = "One Minute Load Average" } metric { name = "load_five" value_threshold = "1.0" title = "Five Minute Load Average" } metric { name = "load_fifteen" value_threshold = "1.0" title = "Fifteen Minute Load Average" } }

/ This group collects the number of running and total processes / collection_group { collect_every = 80 time_threshold = 950 metric { name = "proc_run" value_threshold = "1.0" title = "Total Running Processes" } metric { name = "proc_total" value_threshold = "1.0" title = "Total Processes" } }

/ This collection group grabs the volatile memory metrics every 40 secs and sends them at least every 180 secs. This time_threshold can be increased significantly to reduce unneeded network traffic. / collection_group { collect_every = 40 time_threshold = 180 metric { name = "mem_free" value_threshold = "1024.0" title = "Free Memory" } metric { name = "mem_shared" value_threshold = "1024.0" title = "Shared Memory" } metric { name = "mem_buffers" value_threshold = "1024.0" title = "Memory Buffers" } metric { name = "mem_cached" value_threshold = "1024.0" title = "Cached Memory" } metric { name = "swap_free" value_threshold = "1024.0" title = "Free Swap Space" } }

collection_group { collect_every = 40 time_threshold = 300 metric { name = "bytes_out" value_threshold = 4096 title = "Bytes Sent" } metric { name = "bytes_in" value_threshold = 4096 title = "Bytes Received" } metric { name = "pkts_in" value_threshold = 256 title = "Packets Received" } metric { name = "pkts_out" value_threshold = 256 title = "Packets Sent" } }

/ Different than 2.5.x default since the old config made no sense / collection_group { collect_every = 1800 time_threshold = 3600 metric { name = "disk_total" value_threshold = 1.0 title = "Total Disk Space" } }

collection_group { collect_every = 40 time_threshold = 180 metric { name = "disk_free" value_threshold = 1.0 title = "Disk Space Available" } metric { name = "part_max_used" value_threshold = 1.0 title = "Maximum Disk Space Used" } }