4-way Google Kubernetes Engine Tips for Cold Start Lag

If you use Google Kubernetes Engine for workload execution, it’s likely that you have encountered cold starts

[{"selector":"#anim-a6e10678-49b2-4f8d-b315-b0fd1f8d252d [data-leaf-element=\"true\"]","keyframes":{"transform":["translate3d(36.069749131772724%, 0, 0)","translate3d(0%, 0, 0)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-5820cbd9-67a4-4f5f-816c-eeddb9a5e0c5","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-1b3f789f-0c29-46da-9c1a-f70e8bcbbec1","keyframes":{"transform":["translate3d(0px, -158.52559%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}]

which are delays in application launch caused by workloads assigned to nodes that haven’t hosted the workload

[{"selector":"#anim-420bc6b9-0af0-4f96-a888-11d3e16fc2ae [data-leaf-element=\"true\"]","keyframes":{"transform":["translate3d(-32.42187489333662%, 0, 0)","translate3d(0%, 0, 0)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-21c50ff1-8b79-40c5-b242-04c4db1d312c","keyframes":{"transform":["scale(1)","scale(1.5)","scale(0.95)","scale(1)"],"offset":[0,0.33,0.66,1]},"delay":0,"duration":1450,"easing":"ease-in-out","fill":"both","iterations":1}]

application code are some of the common tasks involved in deploying a containerized application on Kubernetes

[{"selector":"#anim-196428f8-3854-4029-b9e4-e1b3c9def68a [data-leaf-element=\"true\"]","keyframes":{"transform":["translate3d(21.874999829338595%, 0, 0)","translate3d(0%, 0, 0)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-6035bd6d-9c0d-44ef-8e93-7daf9bc1611d","keyframes":{"transform":["translate3d(125.27472%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":1000,"easing":"cubic-bezier(.2, 0, .8, 1)","fill":"both"}] [{"selector":"#anim-8fd8e5f4-f309-4e59-a09b-9b79f0f838b9","keyframes":{"transform":["rotateZ(180deg)","rotateZ(0deg)"]},"delay":0,"duration":1000,"easing":"cubic-bezier(.2, 0, .5, 1)","fill":"forwards"}]

The lack of a pre-existing container image on the new node might result in a much longer initial startup time The pod doesn’t need to start up again since it is already up and heated when a subsequent request comes in

[{"selector":"#anim-2cf497ac-4e39-45eb-9582-d14360444183 [data-leaf-element=\"true\"]","keyframes":{"transform":["translate3d(34.249999904429615%, 0, 0)","translate3d(0%, 0, 0)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-80f5c606-ad6a-42bf-ba1d-bb4718f0e635","keyframes":{"opacity":[0,1]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-344dea65-4636-4673-b226-079a77f96020","keyframes":{"transform":["scale(0.3333333333333333)","scale(1)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"forwards"}]

When pods are being shut down and restarted repeatedly, requests are being sent to fresh, cold pods, which results in a high frequency of cold starts

[{"selector":"#anim-301deb3b-d037-4e1d-a368-29a2927ca4f3 [data-leaf-element=\"true\"]","keyframes":{"transform":["translate3d(-21.874999829338588%, 0, 0)","translate3d(0%, 0, 0)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-864bba4e-59dc-4386-aebb-4cfb50de4fdb","keyframes":{"transform":["translate3d(-122.02797%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-4a07e6c0-e374-4c3c-9d08-c5ee733be097","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-0bd37559-5ff3-4bca-a91c-468112a57189","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

Nevertheless, the warm pool technique may be quite expensive for heavier workloads like AI/ML, particularly on pricey and in-demand GPUs

[{"selector":"#anim-69dcdb89-84c5-4447-b21d-85c694c44078 [data-leaf-element=\"true\"]","keyframes":{"transform":["translate3d(21.874999829338595%, 0, 0)","translate3d(0%, 0, 0)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-71c38d9d-9cc2-43d9-824f-97860d576e43","keyframes":{"transform":["translate3d(-117.32853%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-1a0e8be4-6396-4962-b460-375766a04688","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-46158182-7647-40de-ba1d-d66b66491edc","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

The managed Kubernetes service offered by Google Cloud, Google Kubernetes Engine (GKE), may facilitate the deployment and upkeep of complex containerized workloads

[{"selector":"#anim-6bdc169c-7dce-4d77-b057-6ee024ae7bfa [data-leaf-element=\"true\"]","keyframes":{"transform":["translate3d(32.42187489333662%, 0, 0)","translate3d(0%, 0, 0)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-e73f574f-7ad6-43eb-b942-ea160e149f97","keyframes":{"opacity":[1,1]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-49288d39-c83c-47b6-aee4-3e7522f0681d","keyframes":{"transform":["scale(3)","scale(1)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"forwards"}]

Methods for overcoming the difficulty of chilly starts When using bigger boot drives or local SSDs, use ephemeral storage

[{"selector":"#anim-d1682d2c-0e03-4e79-ad91-0c50f9e514d3 [data-leaf-element=\"true\"]","keyframes":{"transform":["translate3d(-36.069749131772724%, 0, 0)","translate3d(0%, 0, 0)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-b03dad38-8844-4ae8-a307-2c0863acaac0","keyframes":{"transform":["translate3d(137.89063%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":1000,"easing":"cubic-bezier(.2, 0, .8, 1)","fill":"both"}] [{"selector":"#anim-b5afa652-a293-4adc-b822-c0b34650de33","keyframes":{"transform":["rotateZ(180deg)","rotateZ(0deg)"]},"delay":0,"duration":1000,"easing":"cubic-bezier(.2, 0, .5, 1)","fill":"forwards"}]

Google Kubernetes Engine -running apps may be minimized with appropriate design and optimization For more details visit govindhtech.com

[{"selector":"#anim-62432334-4dbe-4d95-a9fa-8efff6880c49 [data-leaf-element=\"true\"]","keyframes":{"transform":["translate3d(-21.874999829338588%, 0, 0)","translate3d(0%, 0, 0)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-c748e650-b1c7-4dd2-9145-49ab9665f64f","keyframes":{"transform":["translate3d(-121.61172%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":1000,"easing":"cubic-bezier(.2, 0, .8, 1)","fill":"both"}] [{"selector":"#anim-5c511f19-2e17-4ad5-90d3-872a085488a2","keyframes":{"transform":["rotateZ(-180deg)","rotateZ(0deg)"]},"delay":0,"duration":1000,"easing":"cubic-bezier(.2, 0, .5, 1)","fill":"forwards"}]