Topics created by IRJ

@IRJ said in Steam Deck - The Linux mobile hardware and OS we have always wanted:

My first thought was why KDE as well. In reality though, no desktop environment is a perfect fit.

Oh for sure. My first hunch would have been to stick with the default that everyone already knows and uses or to go with Deepin because it is so modern and slick.

Don't get me wrong, I'm a fan of KDE and like what they do. It's good to see them keeping a presence in such an important and large scale device.

@Romo will be proud as he's a KDE user.

Scale Down

###################################### ## Save Deployment State (excludes kube,mongo,k8 pods) ###################################### kubectl get deploy -A --no-headers | grep -v -E 'kube|mongo|k8s-api-proxy' > deploy_state_before_scale.txt ###################################### ## Copy Deployment State to GCS Bucket ###################################### gsutil cp deploy_state_before_scale.txt gs://app1 ####################################### ## Scale Deployments to zero ####################################### kubectl get deploy -A --no-headers | grep -v -E 'kube|mongo|k8s-api-proxy' | awk '{print \$1,\$2}' | while read NS DEPLOY; do kubectl scale --replicas=0 deployment/\$DEPLOY -n \$NS; done ####################################### ## Scale Daemons to zero ####################################### kubectl -n <namespace> patch daemonset <name-of-daemon-set> -p '{"spec": {"template": {"spec": {"nodeSelector": {"non-existing": "true"}}}}}' ####################################### ## Turn off Autoscaler on GKE nodepools ####################################### gcloud container clusters update <app1-cluster> --no-enable-autoscaling --region <region> --node-pool <app1nodepool1> gcloud container clusters update <app1-cluster> --no-enable-autoscaling --region <region> --node-pool <app1nodepool2> ####################################### ## Resize Node Pools to zero ####################################### gcloud container clusters update <app1-cluster> --num-nodes 0 --region <region> --node-pool <app1nodepool1> gcloud container clusters update <app1-cluster> --num-nodes 0 --region <region> --node-pool <app1nodepool2>

Scale Up

####################################### ## Resize Node size to 1 for each node pool ####################################### gcloud container clusters update <app1-cluster> --num-nodes 1 --region <region> --node-pool <app1nodepool1> gcloud container clusters update <app1-cluster> --num-nodes 1 --region <region> --node-pool <app1nodepool2> ####################################### ## Turn Autoscaling Back on ####################################### gcloud container clusters update <app1-cluster> --enable-autoscaling --region <region> --node-pool <app1nodepool1> gcloud container clusters update <app1-cluster> --enable-autoscaling --region <region> --node-pool <app1nodepool2> ##################################################### ## Copy Saved Deployment State from GCS bucket ##################################################### gsutil cp gs://<app1>/deploy_state_before_scale.txt . ##################################################### ## Scale deployments using the previously saved state file ##################################################### awk '{print \$1,\$2,\$4}' deploy_state_before_scale.txt | while read NS DEPLOY SCALE; do kubectl scale --replicas=\$SCALE deployment/\$DEPLOY -n \$NS; done ##################################################### ## Scale Daemons back up ##################################################### kubectl -n <namespace> patch daemonset <name-of-daemon-set> --type json -p='[{"op": "remove", "path": "/spec/template/spec/nodeSelector/non-existing"}]'

@irj A Domain Auction site like?, https://www.godaddy.com/en-uk/auctions/domain-auctions

@gjacobse said in Why do we have a downvote limit?:

While I'll defer to @scottalanmiller, it's likely to reduce some 'personal' opinion level votes. I'd rate them at 'vengeance' votes.

^ this.
If left unchecked, things can get nasty and really sour the place.

Also, obligatory xkcd:

https://imgs.xkcd.com/comics/duty_calls.png

So you can also run serverless on K8s. It helps alleviate some of the vendor lock in issues, but doesn't necessarily alleviate cold starts. OpenFaaS and Kubeless are two popular serverless frameworks for K8s. They both allow custom runtimes/templates so you aren't beholdant to a specific vendors security when deploying your apps.

@irj said in Greenfield Kubernetes Architecture and Security:

Also another related question, would you even use kubernetes for databases or would be better to use a hosted service like RDS?

it depends. It's valid either way. However, things like dynamodb can get stupid expensive really quickly, so it's valuable to run those in cluster and just pay for the PVCs used.

@mr-jones said in Ryzen 3900x vs 5600x:

@irj Nice. Since I left that comment I bought a pre-built as a graduation gift to myself as it seems the only way to get anything at the moment. I went with the 3900X / Radeon RX 6700XT combo. I got about a 6 week wait though.

One thing I would consider, if you haven't thought about it already, in relation to your original question is your power consumption. I noticed the 3900X is 105W whereas the 5600X is only 65W.

Unfortunately I found out my motherboard is not backwards compatible so I will be sticking with 5600x for a little while. I woolf eventually like to upgrade to a 5900x when I can find one for MSRP. The upgrade cost from Dell was more than Msrp on 5900x so I will just wait.

I have a 1000watt PSU so CPU power consumption shouldn't matter much for me.

@irj said in Single 32gb DDR4 RAM vs 2x16gb:

I just bought a pre-built desktop from Dell. I opted not to upgrade the memory as it was much more expensive to do through them.

My system is only coming with a single 8gb RAM stick which is obviously unacceptable. The motherboard has 4 RAM slots and supports up to 128GB.

I'm looking at single 32GB sticks vs 2x16. The single sticks are appealing to me because it makes it easier to upgrade in the future. I could eventually have 32x4 to make 128GB.

Now I don't see needing more than 40gb currently which is why I was thinking 32gb would be fine.

Ok that's alot of rambling... My main question is single stick performance vs using RAM in channel. What would you do?

BTDT. I would go with two 32GB.

The future is uncertain and it's much easier to just get what you might need right now. If you can imagine needing 40 GB I'd do 2x32GB right away. Then you have optimum performance and the same memory on both.

@dashrender they reason they had to take it down was because they were basically freely giving out information that they and other companies have to pay for.

Big fan of the that site.

It was fun! 🙂

Sorry, but this is a case of not running the right OS. Wrong tool for the job if you will.

If you need Skype for business, you need Windows. It's as simple as that. Everything else is just a fragile workaround and an exercise in frustration.

Or you could wait until Microsoft have replaced it completely with Teams. Which, if I understand correctly, they've already decided to do.

Or perhaps get one of the Yealink Skype for Business phones. Don't know how it works but I'm guessing it doesn't need anything else.

@scottalanmiller said in Resentment to Purchasing Software - Split From Unrelated Topic on IT Professionals:

@dbeato said in Resentment to Purchasing Software - Split From Unrelated Topic on IT Professionals:

However Windows, Mac OS and any Linux Flavor have their own risks so for me it is about mitigating considering both the usability and risks.

See, this is what I mean. Everything has risks. Risk evaluation involves evaluating how big and common the risks are, what they are likely to do, what the impact is.

If you just say "there is risk", all risk evaluation is useless. IT has to evaluate and consider the risks. Just saying that you have to be afraid of everything because everything has risks and treating it all the same is the same as completely ignoring it. "When everything is a risk, nothing is a risk."

It is understand what you think but saying and doing are two different things. So yes, most things have risks and nothing is perfect but you action it on prevention and being proactive you do no leave it to chance.

Having this issue on another Ubuntu 20.04.

Same error as before, but everything seems to be installed correctly

vagrant plugin list No plugins installed.

Ok HP clarifies and said USB A is not powerful enough to deliver a meaningful charge.

https://store.hp.com/us/en/tech-takes/3-different-ways-to-charge-laptop-without-charger

Bookmarked

That is nice, all the time when updating images is that kernel that gets the most updates.

@black3dynamite said in Ubuntu 20.04 Upgrade / Usage Experience:

@IRJ Did you choose a normal or minimal installation?

Normal because I'm lazy