#Collectd

Collectd

Collectd’s Write HTTP plugin can be used to configure collectd to send data to Metricly. If you need additional information about setting up Collectd, view their wiki. Configuration 1. Copy API Key in Metricly From the top navigation menu, click Integrations. Click the collectd card. Data collection should already be enabled, and a unique API key for your account has already been generated. Copy the API key. 2.

Collectd CPU Metrics

Collected Fully Qualified Name(FQN) Description Statistic Units Min Max Sparse Data Strategy(SDS) BASE CORR UTIL cpu-#.cpu-idle.value Number of jiffies the CPU spent idle. average jiffies 0 none none yes no no cpu-#.cpu-interrupt.value Number of jiffies the CPU spent processing hardware interrupts. average jiffies 0 none none yes no no cpu-#.cpu-nice.value Number of jiffies the CPU spent processing low priority user threads.

Collectd DF Metrics

Collected Fully Qualified Name(FQN) Description Statistic Units Min Max Sparse Data Strategy(SDS) BASE CORR UTIL df-<mount>.df_complex-free.value Free disk space in bytes. average bytes 0 none none yes no no df-<mount>.df_complex-reserved.value Disk space reserved for root user in bytes. average bytes 0 none none yes no no df-<mount>.df_complex-used.value Used disk space in bytes. average bytes 0 none none yes no no Computed Fully Qualified Name(FQN) Description Statistic Units Min Max BASE CORR UTIL metricly.

Collectd Disk Metrics

Collected Fully Qualified Name(FQN) Description Statistic Units Min Max Sparse Data Strategy(SDS) BASE CORR UTIL disk-<dn>.disk_merged.read Number of merged reads per second. average operations/second 0 none none yes no no disk-<dn>.disk_merged.write Number of merged writes per second. average operations/second 0 none none yes no no disk-<dn>.disk_octets.read Bytes read per second. average bytes/second 0 none none yes no no disk-<dn>.

Collectd Interface Metrics

Collected Fully Qualified Name(FQN) Description Statistic Units Min Max Sparse Data Strategy(SDS) BASE CORR UTIL interface-<int>.if_errors.rx Errors per second received. average errors/second 0 none none yes no no interface-<int>.if_errors.tx Errors per second sent. average errors/second 0 none none yes no no interface-<int>.if_octets.rx Bytes per second received. average bytes/second 0 none none yes no no interface-<int>.if_octets.tx Bytes per second sent.

Collectd Load Metrics

Collected Fully Qualified Name(FQN) Description Statistic Units Min Max Sparse Data Strategy(SDS) BASE CORR UTIL load.load.longterm The is the average run queue size over the past 15 minutes. average queued processes 0 none none yes no no load.load.midterm The is the average run queue size over the past 5 minutes. average queued processes 0 none none yes yes no load.

Collectd Memory Metrics

Collected Fully Qualified Name(FQN) Description Statistic Units Min Max Sparse Data Strategy(SDS) BASE CORR UTIL memory.memory-buffered.value Memory being used by buffers; this memory is available to be freed forapplications to use, should they need it. average bytes 0 none none yes no no memory.memory-cached.value Memory being used by caches; this memory is available to be freed forapplications to use, should they need it.

Collectd Policies

Policy name Duration Conditions Category Description Elevated Memory Usage (Collectd) 30 min metricly.collectd.memory.utilizationpercent has an upper baseline deviation INFO Indicates an increase in memory usage above what is considered to be normal. Elevated Process Count 30 min metricly.collectd.processes.total has an upper baseline deviation INFO Indicates that the total number of processes has increased above what is considered to be normal. Elevated Percentage of Blocked Processes 30 min metricly.

Collectd Processes Metrics

Collected Fully Qualified Name(FQN) Description Statistic Units Min Max Sparse Data Strategy(SDS) BASE CORR UTIL processes.fork_rate.value Number of processes per second being forked. average forks/second 0 none none yes no no processes.ps_state-blocked.value Number of processes in a blocked state. A process that is blocked is onethat is waiting for some event, such as a resource becoming available orthe completion of an I/O operation.