The task seems simple. If the file does not exist, create it, otherwise retain the file and its content. In both cases, apply permissions (owner, group, mode).

The most obvious candidate to this task is ansible.builtin.file with state: touch. However, it always appears as changed as the times need to be updated. Rather, once it exists and modes are ok, it should just be ok.

The next candidate likely is ansible.builtin.copy. If passing force: false, it creates the file. However, it does not fix up permission on an already existing file.

It is possible to run containers as a regular user using https://podman.io/ (buzzword: https://github.com/containers/podman#rootless ).

Note that while podman is based on the same OCI standard as docker, there are some differences in the details. For example, healthchecks are specific to docker.

I assume this is what Terragrunt would help to solve. It adds a new layer where you can easily manage complex infrastructure or keep your work DRY.

I had the same issue when trying Microsofts tutorial. Adding host under the rules section worked for me.

Example:

spec:
  ingressClassName: nginx
  rules:
  - host: {your_app}.northeurope.cloudapp.azure.com
    http:
      paths:
      - path: /test(/|$)(.*)
        pathType: Prefix
        backend:
          service:
            name: mockserver-service
            port:
              number: 1234

Got it.

Thanks to https://stackoverflow.com/a/72530077/20003774 , all i had to do was explicitly set an appropriate logging level.

# Use this logger to forward log messages to CloudWatch Logs.
LOG = logging.getLogger(name)

LOG.setLevel(logging.INFO)            # <- this line was missing

When deploying stack, log groups appears. Yay.

I'm taking TechnoWorld Kubernetes Administration class, where we learn to create a Kubernetes cluster from scratch using one Ubuntu EC2 for master node, and two Ubuntu EC2s for worker nodes. The course guides us into installing the different Kubernetes components - containerd, kubeadm, kubelet, kubectl, Weavnet etc. - on each pertinent EC2. We use NGINX as our simple application that are running on our worker node pods. I have the infrastructure set up where I can actually, for example, do a curl from my master node and get the "Welcome to nginx" greeting as a response. So far so good. I even created a Terraform script to set this up. I even setup test Kubernetes LoadBalancer service that's accessible to the outside world, put it's IP:port on the browser and see "Welcome to nginx" proving that the networking (VPC, subnet, security group) is correct.

Here's the traffic flow of the fundamental infrastructure we're trying to build.

Internet-facing AWS ALB(port 80) -> ingress controller LoadBalancer service(port 80:32111 mapping) -> ingress -> ClusterIP service -> EC2 worker with NGINX pod

So a user puts an ALBs DNS name on the browser and gets the "Welcome to NGINX" greeting.

However with my setup, when I put the ALB's DNS name, I get a "This site can't be reached" message.

I set up my internet-facing ALB to serve port 80, forward to a target group on port 32111 (ingress controller port) that has the 2 EC2 worker nodes registered as targets.

Any pointers on how I can debug this seemingly simple/fundamental setup?

TIA

maybe you can go in to a bit more details on the branching strategy you follow for example git flow as that will impact how you setup your CI/CD Pipelines.

For us we have one Pipeline for CD that will build the Development Branch (Git Flow) to deploy to our UAT Environment but automatically skip the Prod environment stage if the branch is called Develop* or Feature*

If the Branch starts with Release/* it will go throe the build stage, the UAT Deployment stage, The Automated Testing stage (Against UAT) and then the Production Stage, see below graphic to illustrate.

What I like on this approach is that you use the exact same code package (artifacts) on UAT and Prod, in your case that would be Dev, UAT and Prod.

In our case the CI Pipeline is a separate Pipeline and uses the same yaml templates for the build but only deploys to the Automated Test environment and is triggered by a Pull Request Policy

Update

Oops, Don't do this. It was pointed out to me that this was wholly insecure as curl http://169.254.169.254/latest/user-data will show you any unprivileged user the private keys. The data gets saved as /run/cloud-init/instance-data.json

Original post

There is no module to make this easier, and there is no argument under system_info (how you add and configure the user) to ease the ability to configure the user's SSH keys. The way I went about this was adding something like this in my main.tf to populate the variable ssh_keys_user

ssh_keys_user = {
  write_files = [
    {
      path        = "/home/ecarroll/.ssh/id_rsa"
      content     = file("./ssh/user/cp-terraform-user-id_rsa")
      owner       = "ecarroll:ecarroll"
      permissions = "0600"
      defer       = true
    },
    {
      path        = "/home/ecarroll/.ssh/id_rsa.pub"
      content     = file("./ssh/user/cp-terraform-user-id_rsa.pub")
      owner       = "ecarroll:ecarroll"
      permissions = "0644"
      defer       = true
    },
    {
      path        = "/home/ecarroll/.ssh/id_ecdsa"
      content     = file("./ssh/user/cp-terraform-user-id_ecdsa")
      owner       = "ecarroll:ecarroll"
      permissions = "0600"
      defer       = true
    },
    {
      path        = "/home/ecarroll/.ssh/id_ecdsa.pub"
      content     = file("./ssh/user/cp-terraform-user-id_ecdsa.pub")
      owner       = "ecarroll:ecarroll"
      permissions = "0644"
      defer       = true
    },
    {
      path        = "/home/ecarroll/.ssh/id_ed25519"
      content     = file("./ssh/user/cp-terraform-user-id_ed25519")
      owner       = "ecarroll:ecarroll"
      permissions = "0600"
      defer       = true
    },
    {
      path        = "/home/ecarroll/.ssh/id_ed25519.pub"
      content     = file("./ssh/user/cp-terraform-user-id_ed25519.pub")
      owner       = "ecarroll:ecarroll"
      permissions = "0644"
      defer       = true
    }
  ]
}

Then what I did was wired it into my cloud-init like this,

write_files:
${ yamlencode( ssh_keys_user.write_files ) }

I generated these files with a Makefile like this,

user/cp-terraform-user-id_ecdsa:
        -mkdir user 2> /dev/null;
        ssh-keygen -C "User key for SSH authentication to repos" -N "" -b 521 -t ecdsa -f "$@";
        touch "$@";
user/cp-terraform-user-id_ed25519:

-mkdir user 2> /dev/null;

ssh-keygen -C "User key for SSH authentication to repos" -N "" -t ed25519 -f "$@";

touch "$@";
user/cp-terraform-user-id_rsa:

-mkdir user 2> /dev/null;

ssh-keygen -C "User key for SSH authentication to repos" -N "" -b 4096 -t rsa -f "$@";

touch "$@";

This works fine. Then I just added the .pub files to BitBucket and GitLab.

I'm not sure what the technical terms are, but whenever I've played pokemon games and successfully threw a ball to catch one, the ball always wiggled three times before clicking and confirming the catch.

Now, I'm playing Pokemon Violet and this weekend I noticed that some of the balls I throw, the animation is slightly different and it only happens to wiggle once before clicking. Of course, I much prefer this, as I've had no pokemon escape between the wiggle and the click. But I'm not sure what I was doing that caused these, I haven't been able to determine a pattern.

Is this a random thing, or can I take specific steps to get more of these one-wiggle catches?

I am having difficulty understanding some differences between an item and a block. This is what I understand so far:

I am mostly confused about whether blocks are also considered items when it is not placed on a coordinate. For example, the block is in a chest, the block is in the player's hand, and the block is in a block with an entity (such as a cobblestone in a furnace). What exactly is considered a block and item?

So let’s say we had a dual type fire ghost, any fire/ghost move will have that 1.5X STAB boost. But when it’s terastallized, let’s say fire, fire moves will have 2X boost, but does ghost moves stay 1.5X or return to 1X?

You could have a scoreboard with the Y value using your execute store (scoreboards are more versatile than data) and have another one with the highest score, using the operation > on scoreboard players operation subcommand

/execute as @a store result score @s Y run data get entity @s Pos[1]
/execute as @a run scoreboard players operation @s Ymax > @s Y

And to display the score, you can use the simple setdisplay:

/scoreboard objectives setdisplay sidebard Ymax

Side note: The starting value for Ymax will be empty, but because of the operation, it will end up with 0 if your Y value is bellow 0 (for some odd reason) rather than whatever negavite value of Y you have.

Recognized!
It's "Sky Jump" from Gottlieb(or a similar flipper with another title released for Italy).

The first image is actually from another Gottlieb flipper called "Psychedelic" and is an editing error of the movie! If you see the entire movie you can see the same image in a scene at minute 5!

My situation: I collect oil from many oil rigs and bring them all to one oil rig, where the huge tanker will come and bring it all to my oil harbour. I have many ships bringing oil to the oil rig with order "Transfer All, No Loading", just the big tanker is allowed to load there.

Nevertheless I just discovered there is oil lost on this rig! It can be 1.650.000 liters and in the next moment it can be 1.626.000 liters. What is happening here? Is there some limitation on oil rigs capacity? Is oil lost when it is not loaded for a while? Do they pump oil back into earth? Simulate leaks? All ships are leaving empty, I am very sure. Only the big tanker gets it. But the big tanker has a long route so it takes a long time until hes back.

No game is unmoddable. However there's a few reasons the barrier-to-entry might be higher for older games:

• Lack of interest. Most people are interested in playing newer games. Older games, especially niche ones, have a much smaller audience.
• Specialized hardware knowledge. Most modern games are playable on x86 (Intel) architecture. However most older consoles had their own CPU architectures. Modding games for these consoles requires knowledge specific to that console. This also means modders may need to write their own reverse-engineering tools.
• Natively compiled code. Nearly all older games were compiled to native machine code, which tends to be a lot of work to reverse engineer. Many (but by no means all) modern games are written in languages like C# or Java, which are compiled to intermediate languages; or have their logic written in a scripting language like Lua. Both cases make the game much easier to reverse engineer.
• Modern gaming engines. Most older games used game engines that were specific to that game, or at best, that development studio. Most modern games use one of only a handful of engines (usually Unreal or Unity). This allows modders of different games to share knowledge and tools with each other, again lowering the barrier-to-entry.

Currently trying to build a capable Assassin for Kalandra league. I'm confused as to how Marks work, as I see there have been some changes. If I have flicker strike linked to multi-strike and poacher's mark, am I applying the mark to everything I hit, so I'll always be hitting something that's marked? Or does it only mark the first thing I hit when I start the chain? I see the marks on the enemies for despair curse that I've got on Blasphemy Support, but not the Poacher's Mark symbol, and am wondering if I"m actually getting benefits of Mark Mastery in this case.

It's for winning the 100x/333x battle:

Get a Crown for Winning the 100x/333x Battle.

If you win the 100x/333x Battle, you can get a crown that will be shown above your name during the Splatfest. It is just a sign of winning the 100x/333x Battle and has no special effect.

https://gamewith.net/splatoon3/article/show/35461

When I'm in a lobby waiting for a match to start I'll often send Booyah https://splatoonwiki.org/wiki/Signal to the other players waiting with me. However, I've never seen a player Booyah back to me. That has me wondering: can other players in the lobby see my signal messages?

On the weapon select screen, I noticed there was a red flag next to my Splattershot. All of my other weapons had a gray flag. What does the red flag next to my Splattershot mean?

By the looks of it, the vanilla game does not have ladders according to https://steamcommunity.com/app/333950/discussions/0/1471967615846949155/ , and a few others such as https://steamcommunity.com/app/333950/discussions/0/1676938384249978624/ .

The post links to https://steamcommunity.com/sharedfiles/filedetails/?id=1121540837 that seems to add a type of ladder, but it doesn't look like a straight vertical kind, but a slightly angled one, which I imagine can still be useful.