Metrics collected by the CloudWatch agent

You can collect metrics from servers by installing the CloudWatch agent on the server. You can install the agent on both Amazon EC2 instances and on-premises servers, and on computers running either Linux, Windows Server, or macOS. If you install the agent on an Amazon EC2 instance, the metrics it collects are in addition to the metrics enabled by default on Amazon EC2 instances.

For information about installing the CloudWatch agent on an instance, see Collect metrics, logs, and traces with the CloudWatch agent.

All metrics discussed in this section are collected directly by the CloudWatch agent.

Metrics collected by the CloudWatch agent on Windows Server instances

On a server running Windows Server, installing the CloudWatch agent enables you to collect the metrics associated with the counters in Windows Performance Monitor. The CloudWatch metric names for these counters are created by putting a space between the object name and the counter name. For example, the % Interrupt Time counter of the Processor object is given the metric name Processor % Interrupt Time in CloudWatch. For more information about Windows Performance Monitor counters, see the Microsoft Windows Server documentation.

The default namespace for metrics collected by the CloudWatch agent is CWAgent, although you can specify a different namespace when you configure the agent.

Metrics collected by the CloudWatch agent on Linux and macOS instances

The following table lists the metrics that you can collect with the CloudWatch agent on Linux servers and macOS computers.

Metric	Description
cpu_time_active	The amount of time that the CPU is active in any capacity. This metric is measured in hundredths of a second. Unit: None
cpu_time_guest	The amount of time that the CPU is running a virtual CPU for a guest operating system. This metric is measured in hundredths of a second. Unit: None
cpu_time_guest_nice	The amount of time that the CPU is running a virtual CPU for a guest operating system, which is low-priority and can be interrupted by other processes. This metric is measured in hundredths of a second. Unit: None
cpu_time_idle	The amount of time that the CPU is idle. This metric is measured in hundredths of a second. Unit: None
cpu_time_iowait	The amount of time that the CPU is waiting for I/O operations to complete. This metric is measured in hundredths of a second. Unit: None
cpu_time_irq	The amount of time that the CPU is servicing interrupts. This metric is measured in hundredths of a second. Unit: None
cpu_time_nice	The amount of time that the CPU is in user mode with low-priority processes, which can easily be interrupted by higher-priority processes. This metric is measured in hundredths of a second. Unit: None
cpu_time_softirq	The amount of time that the CPU is servicing software interrupts. This metric is measured in hundredths of a second. Unit: None
cpu_time_steal	The amount of time that the CPU is in stolen time, which is time spent in other operating systems in a virtualized environment. This metric is measured in hundredths of a second. Unit: None
cpu_time_system	The amount of time that the CPU is in system mode. This metric is measured in hundredths of a second. Unit: None
cpu_time_user	The amount of time that the CPU is in user mode. This metric is measured in hundredths of a second. Unit: None
cpu_usage_active	The percentage of time that the CPU is active in any capacity. Unit: Percent
cpu_usage_guest	The percentage of time that the CPU is running a virtual CPU for a guest operating system. Unit: Percent
cpu_usage_guest_nice	The percentage of time that the CPU is running a virtual CPU for a guest operating system, which is low-priority and can be interrupted by other processes. Unit: Percent
cpu_usage_idle	The percentage of time that the CPU is idle. Unit: Percent
cpu_usage_iowait	The percentage of time that the CPU is waiting for I/O operations to complete. Unit: Percent
cpu_usage_irq	The percentage of time that the CPU is servicing interrupts. Unit: Percent
cpu_usage_nice	The percentage of time that the CPU is in user mode with low-priority processes, which higher-priority processes can easily interrupt. Unit: Percent
cpu_usage_softirq	The percentage of time that the CPU is servicing software interrupts. Unit: Percent
cpu_usage_steal	The percentage of time that the CPU is in stolen time, or time spent in other operating systems in a virtualized environment. Unit: Percent
cpu_usage_system	The percentage of time that the CPU is in system mode. Unit: Percent
cpu_usage_user	The percentage of time that the CPU is in user mode. Unit: Percent
disk_free	Free space on the disks. Unit: Bytes
disk_inodes_free	The number of available index nodes on the disk. Unit: Count
disk_inodes_total	The total number of index nodes reserved on the disk. Unit: Count
disk_inodes_used	The number of used index nodes on the disk. Unit: Count
disk_total	Total space on the disks, including used and free. Unit: Bytes
disk_used	Used space on the disks. Unit: Bytes
disk_used_percent	The percentage of total disk space that is used. Unit: Percent
diskio_iops_in_progress	The number of I/O requests that have been issued to the device driver but have not yet completed. Unit: Count
diskio_io_time	The amount of time that the disk has had I/O requests queued. Unit: Milliseconds The only statistic that should be used for this metric is Sum. Do not use Average.
diskio_reads	The number of disk read operations. Unit: Count The only statistic that should be used for this metric is Sum. Do not use Average.
diskio_read_bytes	The number of bytes read from the disks. Unit: Bytes The only statistic that should be used for this metric is Sum. Do not use Average.
diskio_read_time	The amount of time that read requests have waited on the disks. Multiple read requests waiting at the same time increase the number. For example, if 5 requests all wait for an average of 100 milliseconds, 500 is reported. Unit: Milliseconds The only statistic that should be used for this metric is Sum. Do not use Average.
diskio_writes	The number disk write operations. Unit: Count The only statistic that should be used for this metric is Sum. Do not use Average.
diskio_write_bytes	The number of bytes written to the disks. Unit: Bytes The only statistic that should be used for this metric is Sum. Do not use Average.
diskio_write_time	The amount of time that write requests have waited on the disks. Multiple write requests waiting at the same time increase the number. For example, if 8 requests all wait for an average of 1000 milliseconds, 8000 is reported. Unit: Milliseconds The only statistic that should be used for this metric is Sum. Do not use Average.
ethtool_bw_in_allowance_exceeded	The number of packets queued and/or dropped because the inbound aggregate bandwidth exceeded the maximum for the instance.. This metric is collected only if you have listed it in the ethtool subsection of the metrics_collected section of the CloudWatch agent configuration file. For more information, see Collect network performance metrics Unit: None
ethtool_bw_out_allowance_exceeded	The number of packets queued and/or dropped because the outbound aggregate bandwidth exceeded the maximum for the instance.. This metric is collected only if you have listed it in the ethtool subsection of the metrics_collected section of the CloudWatch agent configuration file. For more information, see Collect network performance metrics Unit: None
ethtool_conntrack_allowance_exceeded	The number of packets dropped because connection tracking exceeded the maximum for the instance and new connections could not be established. This can result in packet loss for traffic to or from the instance. This metric is collected only if you have listed it in the ethtool subsection of the metrics_collected section of the CloudWatch agent configuration file. For more information, see Collect network performance metrics Unit: None
ethtool_linklocal_allowance_exceeded	The number of packets dropped because the PPS of the traffic to local proxy services exceeded the maximum for the network interface. This impacts traffic to the DNS service, the Instance Metadata Service, and the Amazon Time Sync Service. This metric is collected only if you have listed it in the ethtool subsection of the metrics_collected section of the CloudWatch agent configuration file. For more information, see Collect network performance metrics Unit: None
ethtool_pps_allowance_exceeded	The number of packets queued and/or dropped because the bidirectional PPS exceeded the maximum for the instance. This metric is collected only if you have listed it in the ethtool subsection of the metrics_collected section of the CloudWatch agent configuration file. For more information, see Collect network performance metrics. Unit: None
mem_active	The amount of memory that has been used in some way during the last sample period. Unit: Bytes
mem_available	The amount of memory that is available and can be given instantly to processes. Unit: Bytes
mem_available_percent	The percentage of memory that is available and can be given instantly to processes. Unit: Percent
mem_buffered	The amount of memory that is being used for buffers. Unit: Bytes
mem_cached	The amount of memory that is being used for file caches. Unit: Bytes
mem_free	The amount of memory that isn't being used. Unit: Bytes
mem_inactive	The amount of memory that hasn't been used in some way during the last sample period Unit: Bytes
mem_total	The total amount of memory. Unit: Bytes
mem_used	The amount of memory currently in use. Unit: Bytes
mem_used_percent	The percentage of memory currently in use. Unit: Percent
net_bytes_recv	The number of bytes received by the network interface. Unit: Bytes The only statistic that should be used for this metric is Sum. Do not use Average.
net_bytes_sent	The number of bytes sent by the network interface. Unit: Bytes The only statistic that should be used for this metric is Sum. Do not use Average.
net_drop_in	The number of packets received by this network interface that were dropped. Unit: Count The only statistic that should be used for this metric is Sum. Do not use Average.
net_drop_out	The number of packets transmitted by this network interface that were dropped. Unit: Count The only statistic that should be used for this metric is Sum. Do not use Average.
net_err_in	The number of receive errors detected by this network interface. Unit: Count The only statistic that should be used for this metric is Sum. Do not use Average.
net_err_out	The number of transmit errors detected by this network interface. Unit: Count The only statistic that should be used for this metric is Sum. Do not use Average.
net_packets_sent	The number of packets sent by this network interface. Unit: Count The only statistic that should be used for this metric is Sum. Do not use Average.
net_packets_recv	The number of packets received by this network interface. Unit: Count The only statistic that should be used for this metric is Sum. Do not use Average.
netstat_tcp_close	The number of TCP connections with no state. Unit: Count
netstat_tcp_close_wait	The number of TCP connections waiting for a termination request from the client. Unit: Count
netstat_tcp_closing	The number of TCP connections that are waiting for a termination request with acknowledgement from the client. Unit: Count
netstat_tcp_established	The number of TCP connections established. Unit: Count
netstat_tcp_fin_wait1	The number of TCP connections in the FIN_WAIT1 state during the process of closing a connection. Unit: Count
netstat_tcp_fin_wait2	The number of TCP connections in the FIN_WAIT2 state during the process of closing a connection. Unit: Count
netstat_tcp_last_ack	The number of TCP connections waiting for the client to send acknowledgement of the connection termination message. This is the last state right before the connection is closed down. Unit: Count
netstat_tcp_listen	The number of TCP ports currently listening for a connection request. Unit: Count
netstat_tcp_none	The number of TCP connections with inactive clients. Unit: Count
netstat_tcp_syn_sent	The number of TCP connections waiting for a matching connection request after having sent a connection request. Unit: Count
netstat_tcp_syn_recv	The number of TCP connections waiting for connection request acknowledgement after having sent and received a connection request. Unit: Count
netstat_tcp_time_wait	The number of TCP connections currently waiting to ensure that the client received the acknowledgement of its connection termination request. Unit: Count
netstat_udp_socket	The number of current UDP connections. Unit: Count
processes_blocked	The number of processes that are blocked. Unit: Count
processes_dead	The number of processes that are dead, indicated by the X state code on Linux. This metric is not collected on macOS computers. Unit: Count
processes_idle	The number of processes that are idle (sleeping for more than 20 seconds). Available only on FreeBSD instances. Unit: Count
processes_paging	The number of processes that are paging, indicated by the W state code on Linux. This metric is not collected on macOS computers. Unit: Count
processes_running	The number of processes that are running, indicated by the R state code. Unit: Count
processes_sleeping	The number of processes that are sleeping, indicated by the S state code. Unit: Count
processes_stopped	The number of processes that are stopped, indicated by the T state code. Unit: Count
processes_total	The total number of processes on the instance. Unit: Count
processes_total_threads	The total number of threads making up the processes. This metric is available only on Linux instances. This metric is not collected on macOS computers. Unit: Count
processes_wait	The number of processes that are paging, indicated by the W state code on FreeBSD instances. This metric is available only on FreeBSD instances, and is not available on Linux, Windows Server, or macOS instances. Unit: Count
processes_zombies	The number of zombie processes, indicated by the Z state code. Unit: Count
swap_free	The amount of swap space that isn't being used. Unit: Bytes
swap_used	The amount of swap space currently in use. Unit: Bytes
swap_used_percent	The percentage of swap space currently in use. Unit: Percent

Definitions of memory metrics collected by the CloudWatch agent

When the CloudWatch agent collects memory metrics, the source is the host's memory management subsystem. For example, the Linux kernel exposes OS-maintained data in /proc. For memory, the data is in /proc/meminfo.

Each different operating system and architecture has different calculations of the resources that are used by processes. For more information, see the following sections.

During each collection interval, the CloudWatch agent on each instance collects the instance resources and calculates the resources being used by all processes which are running in that instance. This information is reported back to CloudWatch metrics. You can configure the length of the collection interval in the CloudWatch agent configuration file. For more information, see CloudWatch agent configuration file: Agent section.

The following list explains how the memory metrics that the CloudWatch agent collects are defined.

• Active Memory– Memory that is being used by a process. In other words, the memory used by current running apps.

• Available Memory– The memory that can be instantly given to the processes without the system going into swap (also known as virtual memory).

• Buffer Memory– The data area shared by hardware devices or program processes that operate at different speeds and priorities.

• Cached Memory– Stores program instructions and data that are used repeatedly in the operation of programs that the CPU is likely to need next.

• Free Memory– Memory that is not being used at all and is readily available. It is completely free for the system to be used when needed.

• Inactive Memory– Pages that have not been accessed "recently".

• Total Memory– The size of the actual physical memory RAM.

• Used Memory– Memory that is currently in use by programs and processes.

Linux: Metrics collected and calculations used

Metrics collected and units:

• Active (Bytes)

• Available (Bytes)

• Available Percent (Percent)

• Buffered (Bytes)

• Cached (Bytes)

• Free (Bytes)

• Inactive (Bytes)

• Total (Bytes)

• Used (Bytes)

• Used Percent (Percent)

Used memory = Total Memory - Free Memory - Cached memory - Buffer memory

Total memory = Used Memory + Free Memory + Cached memory + Buffer memory

macOS: Metrics collected and calculations used

Metrics collected and units:

• Active (Bytes)

• Available (Bytes)

• Available Percent (Percent)

• Free (Bytes)

• Inactive (Bytes)

• Total (Bytes)

• Used (Bytes)

• Used Percent (Percent)

Available memory = Free Memory + Inactive memory

Used memory = Total Memory - Available memory

Total memory = Available Memory + Used Memory

Windows: Metrics collected

The metrics collected on Windows hosts are listed below. All of these metrics have None for Unit.

• Available bytes

• Cache Faults/sec

• Page Faults/sec

• Pages/sec

There are no calculations used for Windows metrics because the CloudWatch agent parses events from performance counters.

Example: Calculating memory metrics on Linux

As an example, suppose that entering the cat /proc/meminfo command on a Linux host shows the following results:

MemTotal:       3824388 kB
MemFree:         462704 kB
MemAvailable:   2157328 kB
Buffers:         126268 kB
Cached:         1560520 kB
SReclaimable:    289080 kB>

In this example, the CloudWatch agent will collect the following values. All the values that the CloudWatch agent collects and reports are in bytes.

• mem_total: 3916173312 bytes

• mem_available: 2209103872 bytes (MemFree + Cached)

• mem_free: 473808896 bytes

• mem_cached: 1893990400 bytes (cached + SReclaimable

• mem_used: 1419075584 bytes (MemTotal – (MemFree + Buffers + (Cached + SReclaimable)))

• mem_buffered: 129667072 bytes

• mem_available_percent: 56.41%

• mem_used_percent: 36.24% (mem_used / mem_total) * 100