This might be a very obvious question.... but when running on , what should worker_processes be set to? As far as I can tell from googling, nginx isn't cgroup aware, so auto wouldn't work.

Hmm using as my ingress can be a bit annoying at times as it uses IngressRoute object instead of the normal Ingress objects. It means things such as the Jaeger Tracing operator can't create Ingress rules automatically

It's so frustrating that tries to do a chown recursively every time a pod starts up if you have fsGroup set. For a volume with a very large amount of files (eg minio!), it means the pod will take ages to start up.

Finally got my cluster back up. All I wanted to do was upgrade kubernetes, and add a few nodes with Fedora CoreOS, but it exploded in my face. Switched back to CentOS 7 for now.

I bought 4 more nodes for my cluster.... my wife is going to kill me when she finds out 😅

I've been spending the last week going through a training course for . I really want to switch my home cluster from K8S to OpenShift, but it would be problematic as I have very little RAM and CPU for my cluster.

It totals 16GB, and 4 CPU cores across two nodes, and the CPUs aren't great either (G1840T).

I want a couple of quad core i7/i5 NUCs to add to the cluster. If I come across some for a reasonable price, I will definitely be getting them.

I wish FreeIPA had better integration with other tools, such as it would be so nice if or -manager could request certificates from (ie Dogtag), or if ExternalDNS could setup DNS records on FreeIPA's DNS server (BIND).

I've just upgraded the last of my "standalone" docker instances to v2. I'm a little nervous to do it on my as I'm sure it will break a lot of things :(

I set a horizontal pod autoscaler on my sidekiq deployment in my cluster. It's cool seeing the number of pods increase.

Just wiped my two HP Microservers to add them to my home cluster. I used Atomic Host. It's a shame CoreOS isn't stable yet, and isn't compatible with yet.

Hmm really need more nodes for my home cluster. I tried to deploy the Helm chart, but it was pretty much unusable, mainly because my nodes are formed of very crappy Intel Celeron CPUs which are already over worked.

Upgrading my cluster.... I hope this doesn't go badly

Anyone using with , are you able to get the X-Forwarded-For handling working correctly?

I want a few more nodes for my home cluster. At the moment I only have two nodes, each with a crappy CPU (Celeron G1840T). My main requirements were they had to be small, and had to be x86_64, so I found a pair of Acer Veriton N4630G on eBay for very cheap. They currently have 8GB RAM each. I want to add one more node, and also upgrade the CPUs to an i5.

I have way too many VMs running on my servers. I really need to move them to .

Show more
Mastodon@intahnet.co.uk

The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!