|
Also Redhat clarified their plans for CoreOS and said that people were misreading the announcement. It looks like CoreOS lives!(for now at least) https://groups.google.com/forum/m/#!topic/coreos-user/GR4YlF2c1dM
|
# ? Feb 2, 2018 20:22 |
|
|
# ? May 15, 2024 04:14 |
|
freeasinbeer posted:Also Redhat clarified their plans for CoreOS and said that people were misreading the announcement. It looks like CoreOS lives!(for now at least) A couple threads here got confusing for a minute til I realized you changed your forum handle and avatar
|
# ? Feb 3, 2018 02:54 |
|
Docjowles posted:A couple threads here got confusing for a minute til I realized you changed your forum handle and avatar One was foisted on me, other I figured I’d dump a name I made when I was 12.
|
# ? Feb 3, 2018 03:00 |
|
Vulture Culture posted:And yet the only goon I've run across in Hangops or the Rands Leadership Slack rarely posts here anymore I enjoy Rands but I lurk the gently caress out of it.
|
# ? Feb 3, 2018 03:20 |
|
freeasinbeer posted:One was foisted on me, other I figured I’d dump a name I made when I was 12. Fair. The name I used when I first got on the internet (in the 90's, as I am super old) was extremely embarrassing and I'm glad I changed over to just pretending to be a fat old guy instead.
|
# ? Feb 3, 2018 03:32 |
|
Docjowles posted:Fair. The name I used when I first got on the internet (in the 90's, as I am super old) was extremely embarrassing and I'm glad I changed over to just pretending to be a fat old guy instead. You're a step ahead of me...my posts and my were also super embarassing at that time. Every once in awhile I'll come across a post of mine on usenet from the early 90s and I just cringe. (now they're just at a regular level of embarassing)
|
# ? Feb 3, 2018 19:23 |
|
The past few pages of container-talk have convinced me to finally try to learn exactly how they work and start using them myself. Been reading through some documentation while waiting for things to download, but there's something that's confusing me so far: I'm developing on Windows 10 and deploying to Windows Server. Docker says that it doesn't require a VM and can run directly off the underlying OS. But since there's no Windows 10 base image, it looks like I have to download a Windows Server Core image. This makes sense. So is Docker going to run Windows Server in a VM off Windows 10, and then run my containers in the VM? On my target machine, if the Windows Server version matches the docker base image OS version, will it run the containers directly off the underlying OS via the docker engine, or will it still create a Windows Server VM on my Windows Server machine regardless? If it can create VMs when the OS that the container is configured for doesn't match the underlying OS, can I run containers for Linux and Windows on the same box or is it limited to just one OS type per docker engine? Related, could I run containers for Windows Server Core and Windows Server Nano side by side on the same box, even though they are different base images? All very basic stuff I'm sure, but I've not had a chance to get started with containers before now, and I want to make sure I understand things properly and not half-assed.
|
# ? Feb 4, 2018 15:56 |
|
beuges posted:So is Docker going to run Windows Server in a VM off Windows 10, and then run my containers in the VM? No. Windows containers use Hyper-V as a hosting mechanism. Hyper-V actually treats even the host OS as a VM, albeit a very special VM. beuges posted:
Stop thinking in terms of VMs. Containers aren't VMs, they are isolation layers. Windows containers run on the Hyper-V hypervisor to get access to system resources (CPU, memory, disk, etc). The "base image" is more of a set of basic capabilities than it is a full OS. This is why containers start in a few seconds instead of a minute or two -- starting a container doesn't involve booting up a full kernel, it just hooks into the already-running kernel. This is, of course, a massive simplification. beuges posted:If it can create VMs when the OS that the container is configured for doesn't match the underlying OS, can I run containers for Linux and Windows on the same box or is it limited to just one OS type per docker engine? Related, could I run containers for Windows Server Core and Windows Server Nano side by side on the same box, even though they are different base images? Windows can run Windows containers. Linux can run Linux containers. Windows can also run Linux containers, but not at the same time as Windows containers. In the case of Linux containers running on Windows, it actually does use a Linux VM to host the containers. You can run as many different containers from different base images as you want, as long as the OS "flavor" is the same -- Windows or Linux. FWIW, my experience with containers for Windows hasn't been great so far. New Yorp New Yorp fucked around with this message at 16:50 on Feb 4, 2018 |
# ? Feb 4, 2018 16:39 |
|
New Yorp New Yorp posted:Stop thinking in terms of VMs. Containers aren't VMs, they are isolation layers. Windows containers run on the Hyper-V hypervisor to get access to system resources (CPU, memory, disk, etc). The "base image" is more of a set of basic capabilities than it is a full OS. This is why containers start in a few seconds instead of a minute or two -- starting a container doesn't involve booting up a full kernel, it just hooks into the already-running kernel. This is, of course, a massive simplification. Sure, but my understanding/experience of Hyper-V so far has been a means to run VMs, hence my confusion. Also, I was trying to work out how it would handle presenting the Server Core base image to the container when it was actually running on Windows 10, but I guess since they basically share the same kernel for the most part, that makes it a lot easier. This does make things clearer for me though, thanks!
|
# ? Feb 4, 2018 17:15 |
|
beuges posted:Sure, but my understanding/experience of Hyper-V so far has been a means to run VMs, hence my confusion. Also, I was trying to work out how it would handle presenting the Server Core base image to the container when it was actually running on Windows 10, but I guess since they basically share the same kernel for the most part, that makes it a lot easier. No problem! I've been on a Windows containers kick lately, trying to containerize a C# build environment. It's been unpleasant. Linux containers work great, though.
|
# ? Feb 4, 2018 17:38 |
|
Windows container stuff is kind of hard mode. Try spinning up a ghost blog container in a Linux VM first, figure out docker volumes, container networking, passing in env vars first before attempting anything complicated in Windows land.
|
# ? Feb 6, 2018 20:12 |
|
How does everyone do their source control for Kubernetes and interaction with Kubernetes. Dumping a bunch of yaml files into a git repo is with a readme.md explaining what they do is a terrible bad way of doing things. Someone tell me why helm charts should not be used as a deployment mechanism for internally produced applications. How about things like rolling updates. Should I wrap all of the commands associated with doing rolling upgrades in a jenkins task runner that someone can just click on. Should I wrap helm charts with Jenkins? Should I use github as a Helm chart repo?
|
# ? Feb 7, 2018 02:00 |
|
Is there a thing that will scrape my Prometheus endpoints in one secure zone, and then push them to Prometheus gateway on my centralized server in another secure zone so that I'm only allowing a single connection on a single port between the two.
|
# ? Feb 7, 2018 03:11 |
|
Hadlock posted:Is there a thing that will scrape my Prometheus endpoints in one secure zone, and then push them to Prometheus gateway on my centralized server in another secure zone so that I'm only allowing a single connection on a single port between the two. Uh. Maybe abusing pushgateway? Or just have a centralized grafana? Edit: https://prometheus.io/docs/prometheus/latest/federation/ And there seem to be some caveats: https://www.robustperception.io/federation-what-is-it-good-for/ I’d just hook a remote grafana to it depending on your needs. freeasinbeer fucked around with this message at 03:26 on Feb 7, 2018 |
# ? Feb 7, 2018 03:19 |
|
It sounds like Federation is what I need to do? Good link.
|
# ? Feb 7, 2018 04:03 |
|
Methanar posted:How does everyone do their source control for Kubernetes and interaction with Kubernetes. makefiles and yaml Helm is awful, it's a massively overcomplex way to....... write templated yaml, where the template values don't ever change in practice anyway oh well lol edit: ksonnet seems cool, but I haven't used it, nor do I grok it quite. edit edit: Mostly, I think you shouldn't be doing complex poo poo in your resource files, and if you are, you should find or build a better way to do it, or move or abstract that complexity elsewhere. You shouldn't be doing a whole bunch more than `kubectl apply -f foo.yaml` Mao Zedong Thot fucked around with this message at 05:17 on Feb 7, 2018 |
# ? Feb 7, 2018 05:14 |
|
Is there a decent Windows DevOps/Sysadmin/automation for Jackasses course or book any of you can recommend? Small words are a must.
|
# ? Feb 8, 2018 03:10 |
|
Warbird posted:Is there a decent Windows DevOps/Sysadmin/automation for Jackasses course or book any of you can recommend? Small words are a must. Seconding this. I’m a sysadmin at a company that is looking to move to containers when they deploy some of our new website apps. The lead dev is completely familiar with everything on that end, but as a JOAT sysadmin I’m looking to get up to speed. We run IIS/MSSQL for everything currently but they’re shifting to .NET Core so we can run Linux containers. We would be using azure (as it stands currently) for our container environment, so I’m just looking for something to get me started as my list of things to get familiar with keeps growing. I do have a Pluralsight sub at my disposal!
|
# ? Feb 11, 2018 16:25 |
|
Spring Heeled Jack posted:Seconding this. I’m a sysadmin at a company that is looking to move to containers when they deploy some of our new website apps. The lead dev is completely familiar with everything on that end, but as a JOAT sysadmin I’m looking to get up to speed. We run IIS/MSSQL for everything currently but they’re shifting to .NET Core so we can run Linux containers. Starting down a similar path. I just finished The DevOps Handbook and highly recommend it. Gonna grab The Phoenix Project next. I was looking for Windows specific books before the holidays, but it looks like the few Windows specific books are going to be published in the coming months. Get familiar with Powershell if you aren't already as well as some Linux environment. Whether you're on Linux or Windows, Docker underpins everything and Kubernetes seems to coming out on top for orchestration, AKS being managed Kubernetes.
|
# ? Feb 11, 2018 18:09 |
|
So I'm working on a web application, which gets compiled, dockerized, pushed to registry, and launched on the test server by the Gitlab CI. Which was pretty cool to do. I've set it so that the master branch will always be launched publishing on a fixed port. But I would also want feature branches to be launched automatically on separate ports - one port per branch, with newer builds from the same branch (= same image tag) ending up on the same port, so we can try them out in parallel. For this build step, the best I've come up with so far is a script that derives a port number from the branch name (hash, modulo 1000, add 6000, knowing that 6k to 7k are available); while fun, this is not a reliable solution since eventually I'll get a port conflict between two branches and the CI will mysteriously fail. Ideally, I'd want a docker run option that says "publish this container port on the first available host port after $startPort". My next idea was to write a loop that tried to run on 6000, detects if the failure to launch was due to a port conflict, and if so increments the port number and tries again - but this was the point where I thought 'hold on, before I write any more bash lines I should probably check with someone else that I'm not missing a much simpler alternative or running down the wrong rabbit hole'. Am I missing a much simpler alternative or running down the wrong rabbit hole?
|
# ? Feb 16, 2018 15:34 |
|
Docker doesn't have a way to check for an available port bind/mapping before launch, but it can tell you the port mappings for running a container via "docker port ...":code:
Alternatively, you can just bind to port 0 and it'll auto-pick some random-rear end free ephemeral port that you can then query via "docker port" or "docker inspect": code:
|
# ? Feb 16, 2018 16:38 |
|
This thing will reverse proxy the docker port https://github.com/jwilder/nginx-proxy Then you can docker run -p 9999 -e VIRTUAL_HOST=${BRANCH_NAME}.yourdomain.com your-image It's a pain in the rear end to debug though when something goes wrong
|
# ? Feb 16, 2018 18:09 |
|
JehovahsWetness posted:Alternatively, you can just bind to port 0 and it'll auto-pick some random-rear end free ephemeral port that you can then query via "docker port" or "docker inspect" Aaand there's the super simple alternative I didn't know about It's not even random either, it starts at 32768 and looks for the next available port, i.e. exactly what i was about to try manually. Thanks!
|
# ? Feb 16, 2018 22:16 |
|
The NPC posted:Starting down a similar path. I just finished The DevOps Handbook and highly recommend it. Gonna grab The Phoenix Project next. I was looking for Windows specific books before the holidays, but it looks like the few Windows specific books are going to be published in the coming months. Get familiar with Powershell if you aren't already as well as some Linux environment. You can skip the Phoenix Project if you're interested in technical details. It's mostly about Agile project management and does not go into any technical depth at all.
|
# ? Feb 16, 2018 23:23 |
|
you can only skip phoenix project if you have 100% found religion about bottlenecks and local vs systemic optimization
|
# ? Feb 17, 2018 00:47 |
|
You can read the Phoenix Project in like two days so it’s not a huge investment in any case. It’s definitely a book for managers trying to understand why traditional IT service delivery sucks rear end, or people who want to formulate that same argument for their own managers. It won’t teach you a drat thing about containers or kubernetes or Jenkins or infrastructure as code. But it might help you understand or explain why they are cool and good. I think it’s short, entertaining, and insightful enough to be worth a read. It’s literally The Goal retold for IT, if that helps place it.
|
# ? Feb 17, 2018 01:22 |
|
Back when we had an IT reading group, people seemed to really enjoy it, everyone seems to know at least one of the characters in the book in their own organization.
|
# ? Feb 17, 2018 01:41 |
|
That thread owned and I’d love to see it revived. VC was totally carrying it, though, and had a kid which tends to mean the death of things like time to “read” and “think”. I’ve been reading the Google SRE book basically since it came out between caring for two kiddos and somehow still haven’t finished it. I did take a detour to read The Manager’s Path which is very very good, and great fodder for that thread if it does rise from the grave.
|
# ? Feb 17, 2018 01:52 |
|
Docjowles posted:You can read the Phoenix Project in like two days so it’s not a huge investment in any case. It’s definitely a book for managers trying to understand why traditional IT service delivery sucks rear end, or people who want to formulate that same argument for their own managers. It won’t teach you a drat thing about containers or kubernetes or Jenkins or infrastructure as code. But it might help you understand or explain why they are cool and good. I think it’s short, entertaining, and insightful enough to be worth a read. Awesome. Next on my reading list once im done with the end of eternity. For now I'm just waiting for Andrew Harlan to just get hosed in some way. What an rear end in a top hat.
|
# ? Feb 17, 2018 03:15 |
|
My issue with the SRE book is that it was more a collection of stories and blogposts as long form than a book with a clear narrative and guideline. A lot of the material seemed to overlap thematically in a way that seemed redundant. On the other hand, Programming Pearls is similar that way but at least I didn’t think I was getting a paragraph on fixing binary search integer overflow for the fifth time like the SRE book spent talking about monitoring subtopics. I still think Time Management for System Administrators is more important of a read for today’s engineers, honestly. While we learn as engineers how to optimize programs for time and space efficiency, it amazes me to see people fail to spend even 1000th the effort on managing their own time when that’s probably the greatest limiting factor for your programming output in the end. I say this as someone so terrible at time management it’s part of why I decided to never have children. Great goal setting and time management is among the traits of those with long-term career and life success more than knowing more ways to sort data structures than others.
|
# ? Feb 17, 2018 05:09 |
|
necrobobsledder posted:My issue with the SRE book is that it was more a collection of stories and blogposts as long form than a book with a clear narrative and guideline. A lot of the material seemed to overlap thematically in a way that seemed redundant. On the other hand, Programming Pearls is similar that way but at least I didn’t think I was getting a paragraph on fixing binary search integer overflow for the fifth time like the SRE book spent talking about monitoring subtopics. I still think Time Management for System Administrators is more important of a read for today’s engineers, honestly. While we learn as engineers how to optimize programs for time and space efficiency, it amazes me to see people fail to spend even 1000th the effort on managing their own time when that’s probably the greatest limiting factor for your programming output in the end. I say this as someone so terrible at time management it’s part of why I decided to never have children. Great goal setting and time management is among the traits of those with long-term career and life success more than knowing more ways to sort data structures than others.
|
# ? Feb 17, 2018 14:17 |
|
We're about to expand our TeamCity suite to include things like ongoing rolling regressions. We can represent these in TeamCity itself, but this is hitting my sense of smell based on what I read on here awhile ago. I get the impression we should implement most of this rolling regression in a script with a data file in source control. The data file can be adjusted per-commit to ensure that we are testing what's on the HEAD. We've already had issues with somebody changing the test plan in TeamCity and having it -1 all inbound reviews due to a global regression. We had to fix that regression, but none of the commits were targeting that. So I figured instead that being able to pair this QA suite with the current state of the code will be a bulwark against that. The general consensus I've seen is to kind of keep these tools at arm's reach. Use them--yes, but don't try to put everything into them. I don't entirely understand why. I just see more people complaining about having gone all-in and then stepping back versus people making fun of others for not completely committing into putting everything in the CI tool itself.
|
# ? Feb 20, 2018 18:41 |
|
Rocko Bonaparte posted:We're about to expand our TeamCity suite to include things like ongoing rolling regressions. We can represent these in TeamCity itself, but this is hitting my sense of smell based on what I read on here awhile ago. I get the impression we should implement most of this rolling regression in a script with a data file in source control. The data file can be adjusted per-commit to ensure that we are testing what's on the HEAD. We've already had issues with somebody changing the test plan in TeamCity and having it -1 all inbound reviews due to a global regression. We had to fix that regression, but none of the commits were targeting that. So I figured instead that being able to pair this QA suite with the current state of the code will be a bulwark against that.
|
# ? Feb 20, 2018 19:05 |
|
My 0.02 is that if you’ve got everyone on-board with putting everything into a particular CI tool (like TC) then go for it; you can use build chains, snapshot dependencies, etc. to break your build up into logical chunks so that things can be run in isolation/iterated on. That said, this is as much a people problem as it is a technical one. If people are used to being able to do things in increments, even if you give them a way to in TC, they’re just gonna click ‘run’ on whatever and let god sort out the details. I don’t really have an answer for you outside of that. The biggest thing I think helps is making sure that devs understand how to use TC really well.
|
# ? Feb 21, 2018 04:56 |
|
The big thing here is somebody would want to add a step towards a regression, but we don't want unrelated commits that don't even have the matching code to fail. I expect the incompatible code to get through because it has already happened two of three times we did this. We were about to start adding more steps more often, and I don't want it to become a game of us having to get together and turn our keys at the same time.
|
# ? Feb 21, 2018 06:28 |
|
Has anyone tried learning DevOps using small, personal projects? If so, how do you go about them? I've been thinking of getting a DigitalOcean droplet to practice DevOps stuff and tools which should help me with my work and hobbies (been thinking of making a couple of online apps to catalog information related to them). I don't want to be vendor-locked so I want to keep things as open source as possible.
|
# ? Feb 21, 2018 06:35 |
|
The closest I’ve come to that is leveraging the multiple VMs you get with Linux Academy. It’ll do in a pinch for smaller projects, but isn’t perfect. I’ve been debating trying AWS, but haven’t done so yet.
|
# ? Feb 21, 2018 08:38 |
|
Schneider Heim posted:Has anyone tried learning DevOps using small, personal projects? If so, how do you go about them? I've been thinking of getting a DigitalOcean droplet to practice DevOps stuff and tools which should help me with my work and hobbies (been thinking of making a couple of online apps to catalog information related to them). I don't want to be vendor-locked so I want to keep things as open source as possible. I have a small VPS that I got tired of rebuilding when I moved to different providers. So I used puppet to configure it for a bit, tested it locally with vagrant, and now it's a docker compose file with a mix of locally built containers, registry containers, and private registry containers.
|
# ? Feb 21, 2018 15:10 |
|
Schneider Heim posted:Has anyone tried learning DevOps using small, personal projects? If so, how do you go about them? I've been thinking of getting a DigitalOcean droplet to practice DevOps stuff and tools which should help me with my work and hobbies (been thinking of making a couple of online apps to catalog information related to them). I don't want to be vendor-locked so I want to keep things as open source as possible.
|
# ? Feb 21, 2018 17:16 |
|
|
# ? May 15, 2024 04:14 |
|
Schneider Heim posted:Has anyone tried learning DevOps using small, personal projects? If so, how do you go about them? I've been thinking of getting a DigitalOcean droplet to practice DevOps stuff and tools which should help me with my work and hobbies (been thinking of making a couple of online apps to catalog information related to them). I don't want to be vendor-locked so I want to keep things as open source as possible. You can always install Jenkins locally on your machine or spin up a Docker image containing Jenkins and run your personal projects through that. I really wish this was forced behavior for javascript devs at work since they are the worst at "works on my machine" mentality when it comes to jobs failing due to build issues.
|
# ? Feb 21, 2018 17:40 |