Closed taylorsmithgg closed 1 month ago
@kirbsauce Is there anyone on support who can take a look at this?
hi @taylorsmithgg ! Sorry you're having issues. To get you to our support team, would you mind emailing support@kubecost.com or posting on our #support channel in Slack?
@kirbsauce I do not believe there's an active invite link for the slack domain. The current one listed requires @kubecost.com domain.
Apologies, @taylorsmithgg , can you try this -> https://join.slack.com/t/kubecost/shared_invite/zt-1dz4a0bb4-InvSsHr9SQsT_D5PBle2rw
Received the case; keeping this open for now as it may indeed be a bug.
This is expected behavior as automatic implementation of those recommendations via Cluster Controller (1-click and continuous) is only supported on the primary cluster right now.
@kwombach12 this issue can be converted into a feature request and we should document somewhere of this limitation
@rossfisherkc This should be my primary since it's EKS + running in cluster? Is there a config I'm missing?
Apologies, @taylorsmithgg , can you try this -> https://join.slack.com/t/kubecost/shared_invite/zt-1dz4a0bb4-InvSsHr9SQsT_D5PBle2rw
Same result:
grrr. @taylorsmithgg , can you share your email address here and I'll invite you directly? If not, just email me at kirby@kubecost.com. That said, I do see you created the ticket so we should be moving forward there.
@kwombach12 this issue can be converted into a feature request and we should document somewhere of this limitation
I do have a FR for expanding Cluster Controller to the secondary clusters. @bstuder99 Can we add to the docs that cluster controller is only available on the primary cluster today?
grrr. @taylorsmithgg , can you share your email address here and I'll invite you directly? If not, just email me at kirby@kubecost.com. That said, I do see you created the ticket so we should be moving forward there.
The one on the ticket is the correct email
Ok, the article has been updated to be more clear.
@kwombach12 user confirmed that they are indeed using on a single, primary cluster. I'm glad we still got the doc change out of it.
We can treat this as a bug as it stands, sound good?
Is this still an issue?
I want to solve this issue.
Did you fix the issue with Cluster Controller?
The cluster controller can be run directly from the secondary, but currently can't be triggered to run from the primary onto the secondary. So I would say the issue still stands but if desired users can install the controller on secondaries and automate resizing via kubectl directly on the secondaries.
Hi Ajay,
Currently we do not need it.
Thanks,
Vitaly.
On Fri, Feb 16, 2024 at 3:16 PM Ajay Tripathy @.***> wrote:
The cluster controller can be run directly from the secondary, but currently can't be triggered to run from the primary onto the secondary. So I would say the issue still stands but if desired users can install the controller on secondaries and automate resizing via kubectl directly on the secondaries.
— Reply to this email directly, view it on GitHub https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_kubecost_cost-2Danalyzer-2Dhelm-2Dchart_issues_2416-23issuecomment-2D1949270256&d=DwMFaQ&c=udBTRvFvXC5Dhqg7UHpJlPps3mZ3LRxpb6__0PomBTQ&r=GkmpmKrTYVfrylJ4lb1b-YvrconpFuiubK_vRQAMIXlC6FvP1Qgr637pfGzVxv2y&m=Zvhnkz3OtrYwqxOkCGpxXu2pt8huChEfQy4MKJdoAYeG6rXmEEQl8ldsFdzoBFS3&s=CaI1-Dt8Qjwqg7VYv2UPUHPem1ym2h_LKtZIgeHAdB4&e=, or unsubscribe https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_ANXQ5PGWN6CYTQTS5MOJZP3YT65AJAVCNFSM6AAAAAAZ7SYO52VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSNBZGI3TAMRVGY&d=DwMFaQ&c=udBTRvFvXC5Dhqg7UHpJlPps3mZ3LRxpb6__0PomBTQ&r=GkmpmKrTYVfrylJ4lb1b-YvrconpFuiubK_vRQAMIXlC6FvP1Qgr637pfGzVxv2y&m=Zvhnkz3OtrYwqxOkCGpxXu2pt8huChEfQy4MKJdoAYeG6rXmEEQl8ldsFdzoBFS3&s=6ZoPlonq3o6DPuPi8LIUz4lOrJhCvoKx8LPu4ZRGb60&e= . You are receiving this because you commented.Message ID: @.***>
Thank you for reaching out.
We are not provided by kubecost-bug-report.txt.
On Wed, May 1, 2024 at 8:16 PM Taylor Smith @.***> wrote:
Describe the bug On EKS Latest version of both cost analyzer and cluster controller.
This is the only response I'm receiving from the cluster controller:
ERR Error unmarshaling request sizing plan in apply handler: unexpected end of JSON input
To Reproduce Configure cloud provider credentials secret
helm upgrade --install kubecost \ --repo https://kubecost.github.io/cost-analyzer/ https://urldefense.proofpoint.com/v2/url?u=https-3A__kubecost.github.io_cost-2Danalyzer_&d=DwQCaQ&c=udBTRvFvXC5Dhqg7UHpJlPps3mZ3LRxpb6__0PomBTQ&r=GkmpmKrTYVfrylJ4lb1b-YvrconpFuiubK_vRQAMIXlC6FvP1Qgr637pfGzVxv2y&m=UFN70s0_-C-8EecgkrH0fj9WsRi0F6QrKJKAUAbHObozMW7iQHB0TXdkrP6Hqm4w&s=qQwYHu8G2OxO64CYQ4x-mACF9V0KUdm3pK3iC-LfVAA&e= cost-analyzer \ --namespace kubecost --create-namespace \ --set prometheus.nodeExporter.enabled=false \ --set prometheus.serviceAccounts.nodeExporter.create=false \ --set prometheus.kubeStateMetrics.enabled=false \ --set global.prometheus.fqdn=http://prometheus-operated.kube-system.svc:9090 https://urldefense.proofpoint.com/v2/url?u=http-3A__prometheus-2Doperated.kube-2Dsystem.svc-3A9090&d=DwQCaQ&c=udBTRvFvXC5Dhqg7UHpJlPps3mZ3LRxpb6__0PomBTQ&r=GkmpmKrTYVfrylJ4lb1b-YvrconpFuiubK_vRQAMIXlC6FvP1Qgr637pfGzVxv2y&m=UFN70s0_-C-8EecgkrH0fj9WsRi0F6QrKJKAUAbHObozMW7iQHB0TXdkrP6Hqm4w&s=PW9iibjAaELHwDyRmCyLud3CkYhCZj9UOC2JcXOGuN4&e= \ --set global.prometheus.enabled=false \ --set kubecostToken="
" \ --set clusterController.enabled=true \ --set clusterController.kubescaler.resizeAllDefault=true Expected behavior The cluster controller should be able to use continuous cluster sizing.
What impact will this have on your ability to get value out of Kubecost? The current implementation of 1-click resizing has the same effect on the latest version of both the cluster controller and kubecost. At the moment we are having to manually apply all changes.
— Reply to this email directly, view it on GitHub https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_kubecost_features-2Dbugs_issues_75&d=DwMCaQ&c=udBTRvFvXC5Dhqg7UHpJlPps3mZ3LRxpb6__0PomBTQ&r=GkmpmKrTYVfrylJ4lb1b-YvrconpFuiubK_vRQAMIXlC6FvP1Qgr637pfGzVxv2y&m=UFN70s0_-C-8EecgkrH0fj9WsRi0F6QrKJKAUAbHObozMW7iQHB0TXdkrP6Hqm4w&s=qzDLDGdkxOcyswwOGSxvelH9GR-g05deN1Ow2w-3iAs&e=, or unsubscribe https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_ANXQ5PH2GOUYEHD52UJLAV3ZAGAWZAVCNFSM6AAAAABHCYBMH2VHI2DSMVQWIX3LMV43ASLTON2WKOZSGI3TIMZYGAZDENA&d=DwMCaQ&c=udBTRvFvXC5Dhqg7UHpJlPps3mZ3LRxpb6__0PomBTQ&r=GkmpmKrTYVfrylJ4lb1b-YvrconpFuiubK_vRQAMIXlC6FvP1Qgr637pfGzVxv2y&m=UFN70s0_-C-8EecgkrH0fj9WsRi0F6QrKJKAUAbHObozMW7iQHB0TXdkrP6Hqm4w&s=iNF7qhdluwTUKThrlzG2ReWR_vQPgIsCtOOZFrlJTAo&e= . You are receiving this because you commented.Message ID: @.***>
Hello, in an effort to consolidate our bug and feature request tracking, we are deprecating using GitHub to track tickets. If this issue is still outstanding and you have not done so already, please raise a request at https://support.kubecost.com/.
Describe the bug On EKS Latest version of both cost analyzer and cluster controller.
This is the only response I'm receiving from the cluster controller:
To Reproduce Configure cloud provider credentials secret
Expected behavior
The cluster controller should be able to use continuous cluster sizing.
What impact will this have on your ability to get value out of Kubecost? The current implementation of 1-click resizing has the same effect on the latest version of both the cluster controller and kubecost. At the moment we are having to manually apply all changes.
Bug report kubecost-bug-report.txt