Hi all,
Does SAS have a way to stop/start a SAS Viya 4 environment on AWS?
I found some paper referring to "sas-orchestrator" but could not find any doc on how to use it to start/stop.
Any idea?
Thanks
I suggest to start here https://go.documentation.sas.com/doc/en/itopscdc/v_036/itopssrv/n0pwhguy22yhe0n1d7pgi63mf6pb.htm#p05...
There are two cronjobs that will help with this.
I suggest to start here https://go.documentation.sas.com/doc/en/itopscdc/v_036/itopssrv/n0pwhguy22yhe0n1d7pgi63mf6pb.htm#p05...
There are two cronjobs that will help with this.
Hi @BrunoMueller ,
Returning to this question after some time..
Using the SAS supplied sas-stop-all job stops all SAS pods but does not seem to stop the AWS pre-requisite pods and they remain running. For example, see output below of list of pods in the AWS cluster after I ran the sas-stop-all job:
[sas@65ffd157191b /]$ kubectl -n kube-system get pods -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES aws-node-2nmtm 1/1 Running 0 106m 192.168.26.23 ip-192-168-26-23.il-central-1.compute.internal <none> <none> aws-node-nm4kx 1/1 Running 0 107m 192.168.51.33 ip-192-168-51-33.il-central-1.compute.internal <none> <none> cluster-autoscaler-aws-cluster-autoscaler-5d99f7f75c-k8dmh 1/1 Running 0 101m 192.168.47.236 ip-192-168-51-33.il-central-1.compute.internal <none> <none> coredns-7fd98fbdf9-4pprb 1/1 Running 0 112m 192.168.12.32 ip-192-168-51-33.il-central-1.compute.internal <none> <none> coredns-7fd98fbdf9-kk5wk 1/1 Running 0 112m 192.168.50.221 ip-192-168-51-33.il-central-1.compute.internal <none> <none> ebs-csi-controller-67868fdb79-bnvm7 5/5 Running 0 101m 192.168.20.78 ip-192-168-51-33.il-central-1.compute.internal <none> <none> ebs-csi-controller-67868fdb79-dw2mh 5/5 Running 0 101m 192.168.2.59 ip-192-168-26-23.il-central-1.compute.internal <none> <none> ebs-csi-node-6gvvl 3/3 Running 0 101m 192.168.14.30 ip-192-168-26-23.il-central-1.compute.internal <none> <none> ebs-csi-node-7zr8w 3/3 Running 0 101m 192.168.58.137 ip-192-168-51-33.il-central-1.compute.internal <none> <none> kube-proxy-t9tgk 1/1 Running 0 106m 192.168.26.23 ip-192-168-26-23.il-central-1.compute.internal <none> <none> kube-proxy-w25wm 1/1 Running 0 107m 192.168.51.33 ip-192-168-51-33.il-central-1.compute.internal <none> <none> metrics-server-b975dc65f-s7kxt 1/1 Running 0 100m 192.168.5.14 ip-192-168-51-33.il-central-1.compute.internal <none> <none>
I want to shut down all pods including the ones above to save AWS costs. The thing is that if I set the node pools to Desired Count = 0 this will close the node pools and shut down the pods but then I do not know how to start them back again.
There must be some documented way to shut down the AWS EKS cluster completely, at least I hope there is..
Any idea?
Thanks @gwootton
For the sake of anyone interested in this topic here is what I did.
To stop, after the running the "sas-stop-all" job, I modified the Desired Count of the remaining dynamic (autoscaled) node pools that were not shutdown yet to zero and thus forced a shutdown on the remaining running AWS pods.
When I wanted to start back again I changed again from zero to one and then these AWS pods started to execute back again and then I could SAS back again using the "sas-start-all" job and all seems to work fine.
Available on demand!
Missed SAS Innovate Las Vegas? Watch all the action for free! View the keynotes, general sessions and 22 breakouts on demand.