Recommended approach to alert on Pods/Container

[Babar_Qayyum]

Dear All,

What is the recommended and easy approach (I meant the minimum human hours investment) to alert for utilization and resources (CPU, Memory), Container restarted and Container totally went down, etc?

Regards,

Babar

[Mizső]

Hi @Babar_Qayyum 

In this matter I am waiting for this announcment results in the coming releases. Mybe there will be some recommondation and predefined alerts.

Dynatrace expands root cause analysis to Kubernetes with Davis AI | Dynatrace news

Br, Mizső

[Babar_Qayyum]

Hello @Mizső 

Thank you for sharing the blog's link. It seems a great enhancement. Basically, I am looking for recommendations related to infrastructure alerts.

Regards,

Babar

[Mizső]

Hi @Babar_Qayyum ,

Yes I think it will be a imoprtant mile stone for the kubernetes monitoring. I am walking in your shoe. I would also find the applicable thresholds. I have been waiting for these feature since I heard about it.

I have already created a kubernetes dashboards with threshold colouing in the Demo Live. Kubernetes event metrics are required for a small part of it.

☸Kubernetes cluster overview sample with new kubernetes metrics - Dynatrace Demo Live: Demo Live - Dynatrace:
https://demoid.live.dynatrace.com/#dashboard;gtf=-90d%20to%20now;gf=all;id=0817dba0-0d29-42d7-909b-f266a10cce04 

But there will be new predefined Dynatrace kubernetes dashboards form saas 1.251, I guess form managed 1.252.  I am very curious...

Dynatrace SaaS release notes version 1.251 | Dynatrace Docs

Br, Mizső 

[florian_g · Accepted solution]

Hi - I guess the answer here is what my amazing engineering friends are currently testing and polishing. OOTB alerts for common k8s platform issues are just around the corner and I'm already super excited as I'm also testing them currently internally and they're amazing. Let me share some sneak preview here (might change slightly until released), so you see it won't be about  "human-hours", but "human-seconds" soon 🙂 Note, that this is just the first set of alerts, and we will keep expanding them immediately after the first release, with every version 🙂 You'll get alerts for clusters, nodes, namespaces and workloads. All of them can be set tenant-wide (kind of your defaults), and afterward overwritten if certain clusters, namespaces require different alerts or customizations (i.e., thresholds) of them.

[Babar_Qayyum]

Hello @florian_g 

Thank you for the informaiton.

Regards,

Babar

[Mizső]

Hi @florian_g,

Thanks for sharing these information. From now I can't sleep until AG 1.253. 😂

Br, Mizső

[florian_g]

Hi @Mizső - I totally can relate to that. Felt like birthday and all public holidays on the same day, when my amazing engineer friends gave a first end2end demo to me 🤩

PS on expectation mgmt: While internally, we require AG 1.253, I can't promise yet that it will go out with this version (that's why next to the required version, there's still an "in-developement" tag). But, I'm keeping my fingers crossed we can stick with it 🙂 We'll do our best - let's see 😄 

[andrew_smutek]

We're also interested in consuming this capability, so looking forward to it's release.

[florian_g]

👋 great news - can confirm that the mentioned feature will go out in a first version (a subset of alerts) with Dynatrace 1.254 🥳 I'm planning to do a community blog about it to share basics/setup, outlook of upcoming alerts and collect feedback from you all 🙂 Once published, I'll share it here - looking fwd to your feedback 🙂

[florian_g]

here we go: https://community.dynatrace.com/t5/Feedback-channel/Introducing-easy-alerting-on-common-Kubernetes-p... 🙂

[Mizső]

Hi @florian_g,

It is a really great news! Congrats for the DT team!

Have a nice day!

Best regards,

Mizső