Standalone Docker container monitoring - - Alibaba Cloud Documentation Center

This topic explains how to monitor Docker standalone containers to effectively manage and maintain containerized applications.

Prerequisites

Managed Service for Prometheus is activated. Instance billing.
An ECS instance is created. Create and manage an ECS instance by using the ECS console (express version).
Resource Center is activated. Activate Resource Center.

Step 1: Integrate Docker container monitoring

Log on to the Managed Service for Prometheus console. In the left-side navigation pane, click Integration Center.

Click the Docker Container card and follow the on-screen instructions. The following table describes the key parameters.

Parameter	Description
cAdvisor installation mode	cAdvisor (Container Advisor) is an open-source container monitoring tool from Google. It collects resource usage information for containers, such as CPU, memory, network, and disk utilization. Automatic installation: Automatically discovers ECS instances based on your specified filter policy and installs the cAdvisor service on them. Manual installation: You must manually install the cAdvisor service on the ECS instance and open the required port for access within the VPC before you can set up monitoring. Note You must configure the instance's security group to allow inbound traffic on the specified port from the VPC's CIDR block.
Host service discovery mode	Specifies how to discover the ECS instances to monitor.
	Taint label selection	ECS taint label: Excludes ECS instances with the specified taints. This is a reverse selection method.
	Unconditional (select all Linux instances in the current VPC)	-
	Tag selection	ECS instance tag: Selects instances by their ECS instance tags.
	CIDR block selection	ECS CIDR block: Selects instances whose IP addresses fall within the specified CIDR block. If you enter the CIDR block of a VPC, all ECS instances within that VPC are selected.
	Instance ID	ECS instance ID: Separate multiple instance IDs with a comma (,).
cAdvisor service port	Default: 8080.
Metric scrape interval (seconds)	Default: 15.
Automatic security group configuration	When enabled, this feature automatically adds a security group rule if the connectivity check fails.
Custom ECS tag injection	Specify the keys of the ECS tags to inject. The corresponding key-value pairs are then added as labels to Prometheus metrics.

Step 2: View monitoring dashboards

Managed Service for Prometheus provides multiple built-in dashboards to show container monitoring details, such as cluster overview, core components, nodes, and Pods. You can view these dashboards on the Container Service for Kubernetes console, the Application Real-Time Monitoring Service (ARMS) console, or the Managed Service for Prometheus console.

Log on to the Managed Service for Prometheus console. In the left-side navigation pane, click Integration Management.
On the Integration Management page, click the Query Dashboards tab. In the ECS environment section, select a cluster to open its dashboard.

Step 3: Set up alerts

Log on to the Managed Service for Prometheus console. In the left-side navigation pane, click Integration Management.
On the Integration Management page, click the Integrated Environments tab. Select ECS instance, then click the name of the target environment to view its details.
The Component Management tab allows you to configure built-in alert notifications for Prometheus. To view the specific alert rules, select Docker Standalone Container Monitoring from the component type list, and then click the Alert Rule tab. You will find three default built-in alert rules: ContainerDown, HighMemoryUsage, and HighCPUUsage, which are in the Running state by default.
The built-in alert rules generate alert events but do not send alert notifications. To send alert notifications by email or to other platforms, click Edit to configure notification methods.

The alert configuration page also allows you to customize parameters such as the alert threshold, duration, and alert content. For detailed configuration instructions, see Create a Prometheus alert rule.

In the Alert Notification section, select Normal Mode. From the Quick Notification Policy drop-down list, select Do not specify notification policies. Expand Advanced Settings and set the alert check interval to 1 minute.

Metrics

Metric	Type	Description
container_blkio_device_usage_total	COUNTER	Cumulative byte usage for the block I/O device.
container_cpu_cfs_periods_total	COUNTER	Total elapsed enforcement periods.
container_cpu_cfs_throttled_periods_total	COUNTER	Total throttled periods.
container_cpu_cfs_throttled_seconds_total	COUNTER	Total time, in seconds, the container was throttled.
container_cpu_load_average_10s	GAUGE	The container's CPU load average over the last 10 seconds.
container_cpu_schedstat_run_periods_total	COUNTER	Total number of times processes in the cgroup ran on the CPU.
container_cpu_schedstat_runqueue_seconds_total	COUNTER	Cumulative time in seconds that container processes have spent waiting in a run queue.
container_cpu_schedstat_run_seconds_total	COUNTER	Cumulative time in seconds that container processes have been running on the CPU.
container_cpu_system_seconds_total	COUNTER	Cumulative system CPU time consumed, in seconds.
container_cpu_usage_seconds_total	COUNTER	Cumulative CPU time consumed, in seconds.
container_cpu_user_seconds_total	COUNTER	Cumulative user CPU time consumed, in seconds.
container_file_descriptors	GAUGE	Number of open file descriptors for the container.
container_fs_inodes_free	GAUGE	Number of available inodes on the file system.
container_fs_inodes_total	GAUGE	Total inodes on the file system.
container_fs_io_current	GAUGE	Number of I/O operations currently in progress.
container_fs_io_time_seconds_total	COUNTER	Cumulative time in seconds spent on I/O operations.
container_fs_io_time_weighted_seconds_total	COUNTER	Cumulative weighted I/O time in seconds.
container_fs_limit_bytes	GAUGE	Maximum bytes the container can use on the file system.
container_fs_read_seconds_total	COUNTER	Cumulative time in seconds spent on reads.
container_fs_reads_bytes_total	COUNTER	Cumulative bytes read.
container_fs_reads_merged_total	COUNTER	Cumulative number of merged reads.
container_fs_reads_total	COUNTER	Cumulative number of completed reads.
container_fs_sector_reads_total	COUNTER	Cumulative number of completed sector reads.
container_fs_sector_writes_total	COUNTER	Cumulative number of completed sector writes.
container_fs_usage_bytes	GAUGE	Number of bytes used by the container on the file system.
container_fs_write_seconds_total	COUNTER	Cumulative time in seconds spent on writes.
container_fs_writes_bytes_total	COUNTER	Cumulative bytes written.
container_fs_writes_merged_total	COUNTER	Cumulative number of merged writes.
container_fs_writes_total	COUNTER	Cumulative number of completed writes.
container_hugetlb_failcnt	COUNTER	Total times that hugepage usage reached its limit.
container_hugetlb_max_usage_bytes	GAUGE	Maximum recorded hugepage usage in bytes.
container_hugetlb_usage_bytes	GAUGE	Current hugepage usage in bytes.
container_last_seen	GAUGE	Timestamp of when the container was last detected.
container_llc_occupancy_bytes	GAUGE	The container's LLC cache usage, measured by Resource Director Technology (RDT) Memory Bandwidth Monitoring (MBM).
container_memory_bandwidth_bytes	GAUGE	The container's total memory bandwidth usage, measured by Resource Director Technology (RDT) Memory Bandwidth Monitoring (MBM).
container_memory_bandwidth_local_bytes	GAUGE	The container's local memory bandwidth usage, measured by Resource Director Technology (RDT) Memory Bandwidth Monitoring (MBM).
container_memory_cache	GAUGE	Total page cache size in bytes.
container_memory_failcnt	COUNTER	Total times that memory usage reached its limit.
container_memory_failures_total	COUNTER	Cumulative number of memory allocation failures
container_memory_mapped_file	GAUGE	Size of the memory mapped file in bytes.
container_memory_max_usage_bytes	GAUGE	Maximum memory usage in bytes.
container_memory_migrate	GAUGE	Memory migration status.
container_memory_numa_pages	GAUGE	Number of memory pages used per NUMA (Non-Uniform Memory Access) node.
container_memory_rss	GAUGE	The Resident Set Size (RSS) in bytes.
container_memory_swap	GAUGE	The container's current swap memory usage in bytes.
container_memory_usage_bytes	GAUGE	Current memory usage in bytes.
container_memory_working_set_bytes	GAUGE	Current working set size in bytes.
container_network_advance_tcp_stats_total	GAUGE	Advanced TCP connection statistics for the container.
container_network_receive_bytes_total	COUNTER	Cumulative number of bytes received.
container_network_receive_errors_total	COUNTER	Cumulative receive errors.
container_network_receive_packets_dropped_total	COUNTER	Cumulative dropped incoming packets.
container_network_receive_packets_total	COUNTER	Cumulative packets received.
container_network_tcp6_usage_total	GAUGE	The container's TCPv6 connection statistics.
container_network_tcp_usage_total	GAUGE	The container's TCP connection statistics.
container_network_transmit_bytes_total	COUNTER	Cumulative number of bytes transmitted.
container_network_transmit_errors_total	COUNTER	Cumulative transmit errors.
container_network_transmit_packets_dropped_total	COUNTER	Cumulative dropped outgoing packets.
container_network_transmit_packets_total	COUNTER	Cumulative packets transmitted.
container_network_udp6_usage_total	GAUGE	The container's UDPv6 connection statistics.
container_network_udp_usage_total	GAUGE	The container's UDP connection statistics.
container_oom_events_total	COUNTER	Total out of memory (OOM) events for the container.
container_perf_events_scaling_ratio	GAUGE	Scaling ratio for the perf event counter.
container_perf_events_total	COUNTER	Scaled count of perf core events.
container_perf_uncore_events_scaling_ratio	GAUGE	Scaling ratio for the perf uncore event counter. The `PMU` (Performance Monitoring Unit) and `socket` labels specify where the measurement occurred.
container_perf_uncore_events_total	COUNTER	Scaled count of perf uncore events. The `PMU` (Performance Monitoring Unit) and `socket` labels specify where the measurement occurred.
container_processes	GAUGE	Number of processes running in the container.
container_referenced_bytes	GAUGE	Memory, in bytes, referenced by the container since the last measurement. This value is based on the `Referenced` field in `/proc/smaps`.
container_sockets	GAUGE	Number of open sockets for the container.
container_spec_cpu_period	GAUGE	The container's CPU period.
container_spec_cpu_quota	GAUGE	The container's CPU quota.
container_spec_cpu_shares	GAUGE	The container's CPU share.
container_spec_memory_limit_bytes	GAUGE	The container's memory limit in bytes.
container_spec_memory_reservation_limit_bytes	GAUGE	The container's memory reservation limit in bytes.
container_spec_memory_swap_limit_bytes	GAUGE	The container's memory swap limit in bytes.
container_start_time_seconds	GAUGE	Start time of the container as a Unix timestamp.
container_tasks_state	GAUGE	Number of tasks by state: sleeping, running, stopped, uninterruptible, or I/O-wait.
container_threads	GAUGE	Number of threads running in the container.
container_threads_max	GAUGE	Maximum threads allowed in the container.
container_ulimits_soft	GAUGE	The soft ulimit value for the container's root process.
machine_cpu_cache_capacity_bytes	GAUGE	The cache size in bytes assigned to the NUMA node and CPU cores.
machine_cpu_cores	GAUGE	Number of logical CPU cores.
machine_cpu_physical_cores	GAUGE	Number of physical CPU cores.
machine_cpu_sockets	GAUGE	Number of CPU sockets.
machine_dimm_capacity_bytes	GAUGE	Total RAM DIMM capacity, labeled by memory module type.
machine_dimm_count	GAUGE	Number of RAM DIMMs, labeled by memory module type.
machine_memory_bytes	GAUGE	Amount of memory installed on the machine.
machine_swap_bytes	GAUGE	Amount of swap memory available on the machine.
machine_node_distance	GAUGE	The relative distance from one NUMA node to another.
machine_node_hugepages_count	GAUGE	Number of hugepages assigned to the NUMA node.
machine_node_memory_capacity_bytes	GAUGE	Memory assigned to the NUMA node, in bytes.
machine_nvm_avg_power_budget_watts	GAUGE	Average power budget of the NVM (Non-Volatile Memory) in watts.
machine_nvm_capacity	GAUGE	The capacity of the NVM (Non-Volatile Memory).
machine_thread_siblings_count	GAUGE	Number of CPU thread siblings.

References

How to Build an Efficient and Stable Docker Container Monitoring System