Kubernetes/Requests and Limits, units - Revision history

Pio2pio: /* Set requests and limits */

2022-03-30T20:39:13Z

Set requests and limits

← Older revision		Revision as of 22:39, 30 March 2022
Line 94:		Line 94:
	;Has the cluster enough capacity? Finding the overcommit of the cluster.		;Has the cluster enough capacity? Finding the overcommit of the cluster.
	<source lang=bash>		<source lang=bash>
	## % memory overcommitted of the cluster		# % memory overcommitted of the cluster
	100 * sum(kube_pod_container_resource_limits{container!="",resource="memory"} ) / sum(kube_node_status_capacity_memory_bytes)		100 * sum(kube_pod_container_resource_limits{container!="",resource="memory"} ) / sum(kube_node_status_capacity_memory_bytes)

	## % CPU overcommitted of the cluster		# % CPU overcommitted of the cluster
	100 * sum(kube_pod_container_resource_limits{container!="",resource="cpu"} ) / sum(kube_node_status_capacity_cpu_cores)		100 * sum(kube_pod_container_resource_limits{container!="",resource="cpu"} ) / sum(kube_node_status_capacity_cpu_cores)
	</source>		</source>

Pio2pio: /* Set requests and limits */

2022-03-30T20:38:58Z

Set requests and limits

← Older revision		Revision as of 22:38, 30 March 2022
Line 105:		Line 105:
	It is also important to check the overcommit per node. An example of node overcommit can be a pod with a request of 2 CPUs and a limit of 8. That pod can be scheduled in a node with 4 cores, but as the pod has 8 cores as the limit, the overcommit in that node would be 8 – 4 = 4 cores.		It is also important to check the overcommit per node. An example of node overcommit can be a pod with a request of 2 CPUs and a limit of 8. That pod can be scheduled in a node with 4 cores, but as the pod has 8 cores as the limit, the overcommit in that node would be 8 – 4 = 4 cores.
	<source lang=bash>		<source lang=bash>
	## % memory overcommitted of the node		# % memory overcommitted of the node
	sum by (node)(kube_pod_container_resource_limits{container!="",resource="memory"} ) / sum by (node)(kube_node_status_capacity_memory_bytes)		sum by (node)(kube_pod_container_resource_limits{container!="",resource="memory"} ) / sum by (node)(kube_node_status_capacity_memory_bytes)

	## % CPU overcommitted of the node		# % CPU overcommitted of the node
	sum by (node)(kube_pod_container_resource_limits{container!="",resource="cpu"} ) / sum by (node)(kube_node_status_capacity_cpu_cores)		sum by (node)(kube_pod_container_resource_limits{container!="",resource="cpu"} ) / sum by (node)(kube_node_status_capacity_cpu_cores)
	</source>		</source>

Pio2pio: /* Set requests and limits */

2022-03-30T20:38:42Z

Set requests and limits

← Older revision		Revision as of 22:38, 30 March 2022
Line 53:		Line 53:
	= Set requests and limits =		= Set requests and limits =
	* [https://sysdig.com/blog/kubernetes-resource-limits/ PromQL sources]		* [https://sysdig.com/blog/kubernetes-resource-limits/ PromQL sources]


			;Find the containers without Kubernetes resource limits
	<source lang=bash>		<source lang=bash>
	~~# Find the containers without Kubernetes resource limits~~		# Containers without CPU limits by namespace

	## Containers without CPU limits by namespace
	sum by (namespace)(count by (namespace,pod,container)(kube_pod_container_info{container!=""}) unless sum by (namespace,pod,container)(kube_pod_container_resource_limits{resource="cpu"}))		sum by (namespace)(count by (namespace,pod,container)(kube_pod_container_info{container!=""}) unless sum by (namespace,pod,container)(kube_pod_container_resource_limits{resource="cpu"}))

	## Containers without memory limits by namespace		# Containers without memory limits by namespace
	sum by (namespace)(count by (namespace,pod,container)(kube_pod_container_info{container!=""}) unless sum by (namespace,pod,container)(kube_pod_container_resource_limits{resource="memory"}))		sum by (namespace)(count by (namespace,pod,container)(kube_pod_container_info{container!=""}) unless sum by (namespace,pod,container)(kube_pod_container_resource_limits{resource="memory"}))

	## Top 10 containers without CPU limits, using more CPU		# Top 10 containers without CPU limits, using more CPU
	topk(10,sum by (namespace,pod,container)(rate(container_cpu_usage_seconds_total{container!=""}[5m])) unless sum by (namespace,pod,container)(kube_pod_container_resource_limits{resource="cpu"}))		topk(10,sum by (namespace,pod,container)(rate(container_cpu_usage_seconds_total{container!=""}[5m])) unless sum by (namespace,pod,container)(kube_pod_container_resource_limits{resource="cpu"}))

	## Top 10 containers without memory limits, using more memory		# Top 10 containers without memory limits, using more memory
	topk(10,sum by (namespace,pod,container)(container_memory_usage_bytes{container!=""}) unless sum by (namespace,pod,container)(kube_pod_container_resource_limits{resource="memory"}))		topk(10,sum by (namespace,pod,container)(container_memory_usage_bytes{container!=""}) unless sum by (namespace,pod,container)(kube_pod_container_resource_limits{resource="memory"}))
			</source>


	# Detecting containers with very tight CPU limits could cause CPU throttling		;Detecting containers with very tight CPU limits could cause CPU throttling
	## CPU usage is close to its limits		<source lang=bash>
			# CPU usage is close to its limits
	(sum by (namespace,pod,container)(rate(container_cpu_usage_seconds_total{container!=""}[5m])) / sum by (namespace,pod,container)(kube_pod_container_resource_limits{resource="cpu"})) > 0.8		(sum by (namespace,pod,container)(rate(container_cpu_usage_seconds_total{container!=""}[5m])) / sum by (namespace,pod,container)(kube_pod_container_resource_limits{resource="cpu"})) > 0.8

	## Memory usage is close to its limits - could cause OOMKilled		# Memory usage is close to its limits - could cause OOMKilled
	(sum by (namespace,pod,container)(container_memory_usage_bytes{container!=""}) / sum by (namespace,pod,container)(kube_pod_container_resource_limits{resource="memory"})) > 0.8		(sum by (namespace,pod,container)(container_memory_usage_bytes{container!=""}) / sum by (namespace,pod,container)(kube_pod_container_resource_limits{resource="memory"})) > 0.8
			</source>


	# How to set the right Kubernetes resource limits?		;How to set the right Kubernetes resource limits?
	# Conservative - We will select the value of the container that consumed the most in each moment. If we set the limit to that value, the containers won’t run out of resources.		Conservative - We will select the value of the container that consumed the most in each moment. If we set the limit to that value, the containers won’t run out of resources.
			<source lang=bash>
	## Finding the right CPU limit, with the conservative strategy		# Finding the right CPU limit, with the conservative strategy
	max by (namespace,owner_name,container)((rate(container_cpu_usage_seconds_total{container!="POD",container!=""}[5m])) * on(namespace,pod) group_left(owner_name) avg by (namespace,pod,owner_name)(kube_pod_owner{owner_kind=~"DaemonSet\|StatefulSet\|Deployment"}))		max by (namespace,owner_name,container)((rate(container_cpu_usage_seconds_total{container!="POD",container!=""}[5m])) * on(namespace,pod) group_left(owner_name) avg by (namespace,pod,owner_name)(kube_pod_owner{owner_kind=~"DaemonSet\|StatefulSet\|Deployment"}))

			# Finding the right memory limit, with the conservative strategy
	## Finding the right memory limit, with the conservative strategy
	max by (namespace,owner_name,container)((container_memory_usage_bytes{container!="POD",container!=""}) * on(namespace,pod) group_left(owner_name) avg by (namespace,pod,owner_name)(kube_pod_owner{owner_kind=~"DaemonSet\|StatefulSet\|Deployment"}))		max by (namespace,owner_name,container)((container_memory_usage_bytes{container!="POD",container!=""}) * on(namespace,pod) group_left(owner_name) avg by (namespace,pod,owner_name)(kube_pod_owner{owner_kind=~"DaemonSet\|StatefulSet\|Deployment"}))
			</source>

	~~# Has the cluster enough capacity?~~
	~~## Finding the overcommit of the cluster~~

			;Has the cluster enough capacity? Finding the overcommit of the cluster.
			<source lang=bash>
	## % memory overcommitted of the cluster		## % memory overcommitted of the cluster
	100 * sum(kube_pod_container_resource_limits{container!="",resource="memory"} ) / sum(kube_node_status_capacity_memory_bytes)		100 * sum(kube_pod_container_resource_limits{container!="",resource="memory"} ) / sum(kube_node_status_capacity_memory_bytes)
Line 95:		Line 99:
	## % CPU overcommitted of the cluster		## % CPU overcommitted of the cluster
	100 * sum(kube_pod_container_resource_limits{container!="",resource="cpu"} ) / sum(kube_node_status_capacity_cpu_cores)		100 * sum(kube_pod_container_resource_limits{container!="",resource="cpu"} ) / sum(kube_node_status_capacity_cpu_cores)
			</source>


	# Finding the overcommit of the node		;Finding the overcommit of the node
			It is also important to check the overcommit per node. An example of node overcommit can be a pod with a request of 2 CPUs and a limit of 8. That pod can be scheduled in a node with 4 cores, but as the pod has 8 cores as the limit, the overcommit in that node would be 8 – 4 = 4 cores.
	# It is also important to check the overcommit per node. An example of node overcommit can be a pod with a request of 2 CPUs and a limit of 8. That pod can be scheduled in a node with 4 cores, but as the pod has 8 cores as the limit, the overcommit in that node would be 8 – 4 = 4 cores.		<source lang=bash>

	## % memory overcommitted of the node		## % memory overcommitted of the node
	sum by (node)(kube_pod_container_resource_limits{container!="",resource="memory"} ) / sum by (node)(kube_node_status_capacity_memory_bytes)		sum by (node)(kube_pod_container_resource_limits{container!="",resource="memory"} ) / sum by (node)(kube_node_status_capacity_memory_bytes)

Pio2pio at 20:34, 30 March 2022

2022-03-30T20:34:47Z

@@ Line 1: / Line 1: @@
 = Limits and requests =
-;CPU
+;CPU - [https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/#meaning-of-cpu Meaning of CPU]
 Limits and requests for CPU resources are measured in cpu units. One cpu, in Kubernetes, is equivalent to 1 vCPU/Core for cloud providers and 1 hyperthread on bare-metal Intel processors.
@@ Line 49: / Line 48: @@
 128974848≈129e6≈129M       # Fixed-point (metric)
 128974848≈129e6≈129M≈123Mi # Binary
 </source>

Pio2pio: Pio2pio moved page Kubernetes/Requests and Limits to Kubernetes/Requests and Limits, units without leaving a redirect

2021-12-08T17:34:30Z

Pio2pio moved page Kubernetes/Requests and Limits to Kubernetes/Requests and Limits, units without leaving a redirect

← Older revision	Revision as of 19:34, 8 December 2021
(No difference)

Pio2pio: /* Limits and requests */

2021-12-08T17:33:42Z

Limits and requests

← Older revision		Revision as of 19:33, 8 December 2021
Line 6:		Line 6:

	;Memory		;Memory
	Limits and requests for memory are measured in '''bytes'''. You can express memory as a plain integer or as a fixed-point number using one of these suffixes: E, P, T, G, M, k, m (millis). You can also use the power-of-two equivalents: Ei, Pi, Ti, Gi, Mi, Ki.		Limits and requests for memory are measured in '''bytes'''. You can express memory as a plain integer (bytes) or as a fixed-point number using one of these suffixes: E, P, T, G, M, k, m (millis). You can also use the power-of-two equivalents: Ei, Pi, Ti, Gi, Mi, Ki.

	For example, the following represent roughly the same value:		For example, the following represent roughly the same value:

Pio2pio: /* Limits and requests */

2021-12-08T17:33:08Z

Limits and requests

← Older revision		Revision as of 19:33, 8 December 2021
Line 5:		Line 5:


	;~~Meaning of memory~~		;Memory
	Limits and requests for memory are measured in bytes. You can express memory as a plain integer or as a fixed-point number using one of these suffixes: E, P, T, G, M, k, m (millis). You can also use the power-of-two equivalents: Ei, Pi, Ti, Gi, Mi, Ki.		Limits and requests for memory are measured in '''bytes'''. You can express memory as a plain integer or as a fixed-point number using one of these suffixes: E, P, T, G, M, k, m (millis). You can also use the power-of-two equivalents: Ei, Pi, Ti, Gi, Mi, Ki.

	For example, the following represent roughly the same value:		For example, the following represent roughly the same value:
	<code>128974848, 129e6, 129M, 128974848000m, 123Mi</code>		<code>128974848, 129e6, 129M, 128974848000m, 123Mi</code>


	= CPU units =		= CPU units =

Pio2pio at 17:27, 8 December 2021

2021-12-08T17:27:21Z

← Older revision		Revision as of 19:27, 8 December 2021
Line 1:		Line 1:
	;[https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/#meaning-of-cpu Meaning of CPU]		;[https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/#meaning-of-cpu Meaning of CPU]
			= Limits and requests =
			;CPU
	Limits and requests for CPU resources are measured in cpu units. One cpu, in Kubernetes, is equivalent to 1 vCPU/Core for cloud providers and 1 hyperthread on bare-metal Intel processors.		Limits and requests for CPU resources are measured in cpu units. One cpu, in Kubernetes, is equivalent to 1 vCPU/Core for cloud providers and 1 hyperthread on bare-metal Intel processors.

Line 11:		Line 12:


	;CPU units		= CPU units =
	The CPU resource is measured in CPU units. <code>1 CPU</code>, in Kubernetes, is equivalent to:		The CPU resource is measured in CPU units. <code>1 CPU</code>, in Kubernetes, is equivalent to:
	* <code>1 vCPU/Core</code> for cloud providers (1 AWS vCPU, 1 GCP Core, 1 Azure vCore)		* <code>1 vCPU/Core</code> for cloud providers (1 AWS vCPU, 1 GCP Core, 1 Azure vCore)

Pio2pio: /* Units */

2021-12-08T17:22:01Z

Units

← Older revision		Revision as of 19:22, 8 December 2021
Line 22:		Line 22:


	= Units =		= Memory Units =
	;Fixed-Point Suffix		;Fixed-Point Suffix
	* E = Exabyte. 1E = 1,000,000,000,000,000,000 bytes		* E = Exabyte. 1E = 1,000,000,000,000,000,000 bytes

Pio2pio at 17:20, 8 December 2021

2021-12-08T17:20:59Z

← Older revision		Revision as of 19:20, 8 December 2021
Line 5:		Line 5:

	;Meaning of memory		;Meaning of memory

	Limits and requests for memory are measured in bytes. You can express memory as a plain integer or as a fixed-point number using one of these suffixes: E, P, T, G, M, k, m (millis). You can also use the power-of-two equivalents: Ei, Pi, Ti, Gi, Mi, Ki.		Limits and requests for memory are measured in bytes. You can express memory as a plain integer or as a fixed-point number using one of these suffixes: E, P, T, G, M, k, m (millis). You can also use the power-of-two equivalents: Ei, Pi, Ti, Gi, Mi, Ki.

Line 13:		Line 12:

	;CPU units		;CPU units
			The CPU resource is measured in CPU units. <code>1 CPU</code>, in Kubernetes, is equivalent to:
			* <code>1 vCPU/Core</code> for cloud providers (1 AWS vCPU, 1 GCP Core, 1 Azure vCore)
			* <code>1 Hyperthread</code> on a bare-metal Intel processor with Hyperthreading

	~~The CPU resource is measured in CPU units. One CPU, in Kubernetes, is equivalent to:~~

	* 1 AWS vCPU
	* 1 GCP Core
	* 1 Azure vCore
	* 1 Hyperthread on a bare-metal Intel processor with Hyperthreading

	Fractional values are allowed. A Container that requests 0.5 CPU is guaranteed half as much CPU as a Container that requests 1 CPU. You can use the suffix m to mean milli. For example 100m CPU, 100 milliCPU, and 0.1 CPU are all the same. Precision finer than 1m is not allowed.		Fractional values are allowed. A Container that requests 0.5 CPU is guaranteed half as much CPU as a Container that requests 1 CPU. You can use the suffix m to mean milli. For example 100m CPU, 100 milliCPU, and 0.1 CPU are all the same. Precision finer than 1m is not allowed.