Scaling a Multi-Tenant k8s Cluster in a Telco Pablo Moncada October 28, 2020 About MasMovil group ● 4th telecom company in Spain ● Provides voice and broadband services to +12M customers ● Several

TIP: To change picture:Right click on image > Replace image > Select file 3 Requirements for Scaling Up TIP: To change picture:Right click on image > Replace image > Select file ❏ Secure Network

--cluster eni-cluster -n ng-25560078 -N 2 [ℹ] scaling nodegroup stack "eksctl-test-cluster-nodegroup-ng- 25560078" in cluster eksctl-test-cluster-cluster [ℹ] scaling nodegroup, desired capacity from 0 to 2 --cluster eni-cluster -n ng-25560078 -N 2 [ℹ] scaling nodegroup stack "eksctl-test-cluster-nodegroup-ng- 25560078" in cluster eksctl-test-cluster-cluster [ℹ] scaling nodegroup, desired capacity from 0 to 2 --cluster eni-cluster -n ng-25560078 -N 2 [ℹ] scaling nodegroup stack "eksctl-test-cluster-nodegroup-ng- 25560078" in cluster eksctl-test-cluster-cluster [ℹ] scaling nodegroup, desired capacity from 0 to 2

--cluster test-cluster -n ng-25560078 -N 2 [ℹ] scaling nodegroup stack "eksctl-test-cluster-nodegroup-ng- 25560078" in cluster eksctl-test-cluster-cluster [ℹ] scaling nodegroup, desired capacity from 0 to 2 --cluster eni-cluster -n ng-25560078 -N 2 [ℹ] scaling nodegroup stack "eksctl-test-cluster-nodegroup-ng- 25560078" in cluster eksctl-test-cluster-cluster [ℹ] scaling nodegroup, desired capacity from 0 to 2

enable the former, run: tuned-adm profile network-latency Set CPU governor to performance The CPU scaling up and down can impact latency tests and lead to sub-optimal performance. To achieve maximum consistent performance. Set the CPU governor to performance: for CPU in /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor; do echo performance > $CPU done Stop irqbalance and pin the NIC interrupts to specific increased with bpf-policy-map-max in the Cilium ConfigMap. The first mark represents the action of scaling up the deployments, the second mark when 50000 pods were in ready state. It is expected to see the

enable the former, run: tuned-adm profile network-latency Set CPU governor to performance The CPU scaling up and down can impact latency tests and lead to sub-optimal performance. To achieve maximum consistent performance. Set the CPU governor to performance: for CPU in /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor; do echo performance > $CPU done Stop irqbalance and pin the NIC interrupts to specific increased with bpf-policy-map-max in the Cilium ConfigMap. The first mark represents the action of scaling up the deployments, the second mark when 50000 pods were in ready state. It is expected to see the

to a limita�on in the load-balancing implementa�on, constant churn in the load balancing table by scaling the number of service backends up and down can reduce the maximum number of supported service backends

increase with bpf-policy-map-max in the Cilium ConfigMap. The first mark represents the action of scaling up the deployments, the second mark when 50000 pods were in ready state. It is expected to see the

increased with bpf-policy-map-max in the Cilium ConfigMap. The first mark represents the action of scaling up the deployments, the second mark when 50000 pods were in ready state. It is expected to see the