AWS CloudWatch Alarms in Slack

https://towardsaws.com/aws-cloudwatch-alarms-in-slack-f2691176d1cf

https://redd.it/oeb821
@r_devops

Best Image Resizing SAAS

What is the best image resizing SAAS Solution according to you?

https://redd.it/oe4s12
@r_devops

Should i as devops replace Dell XPS for Macbook Pro with Intel?

I have Dell precision 5510 which basically is xps 9550 but with quadro
15" 4k touch screen
i7 6820 hq
32gb ram
M2 ssd + sata ssd
quadro m1000m

With Kubuntu on board it does well whathever i throw on it. Linux support is really good, sotware updates contain even bios/uefi! Kubuntu is able to scale 1.25 which is perfect for me.
However i have complaints:
- works with full speed only with cable (serious throttling without it)
- batterry life is pure garbage - > ~40-60min of work
- external 4k screen has low fps (definetly not 60hz)
- i have to restart it when i change scalling (switch to other monitors)
- i have to connect monitors AFTER os is loaded, otherwise some weird things happenin

My company buys new laptops for workers which become their own after some time. Most of coworkers take M1 macbooks. I see it as oppotunity to switch my current gear for something better. Im looking at macbooks now because of their 4k handling, great battery life and no throttling even without power cord. M1 is out of the question for now - still some problems with some tools and no possibility to virtualize windows which i sometimes do. So the only option is MBP 16 2020 with intel cpu and scissor keyboard.

I still consider some dell/thinkpad with linux. But im afraid that i woul gain very little ro none comparing to my current laptop. Mybe systems76 would be better at some points if they would sell them in europe. But they dont.

So what is your experience and reccomendation? My stack is bash, python, docker, docker compose, virtual machines and soon terraform and kubernetes.

https://redd.it/oeh52y
@r_devops

What languages are you typically using use in your devops professional work?

Here let's avoid things like yaml of only configuration files.

View Poll

https://redd.it/oekiq8
@r_devops

Project and/or grind leetcode or certs for DevOps

Had DevOps experience though mostly was a sysadmin position with lots of IaC. I'm currently job searching and was wondering, with less than a year's experience, would it be good for me to just continue grinding leetcode and/or creating meaningful projects (Planning on creating 3 - 4 just for a portfolio) to show case to an employer or just getting a cert in something like AWS Certified Solutions Architect, tack it on the resume/portfolio and then apply?

Would love to hear what your guys thoughts and open to any questions.

https://redd.it/oejnrc
@r_devops

Question about downtime of blue-green deployment

I have read some articles about blue-green and most of those are mentioned it as a "zero downtime" deployment strategy.

But I'm confused on the switch traffic part. It should have downtime when we switch the whole route from version 1 to version 2, right?

I'm using k8s to do my deployment. I face downtime(about 1s to 2s) when I change the selector in service.yaml

Here are my deployment steps:

1. currently running version 1.0.0
2. apply version 2.0.0 deployment.yaml
3. wait the pod running normally and passed health check
4. edit service.yaml, change selector from version 1.0.0 to version 2.0.0

* but here I want to mention that I'm using Lens IDE to edit the service.yaml, I watched some tutorial that teach us use patch command, something like `kubectl patch service my-app -p '{"spec":{"selector":{"version":"v2.0.0"}}}'` to do it. I don't know if it is different behind the Lens IDE.

So finally I want to ask is blue-green really a zero downtime deployment strategy? I know I can use Canary deployment to achieve "real zero downtime". But how can I achieve zero downtime in blue-green only?

Thank you very much~

https://redd.it/oeib4l
@r_devops

Terraform EKS Ubuntu AMI

I'm new to DevOps, and I have been trying to use the Ubuntu image as a node in the AWS EKS cluster in Terraform. But the terraform doesn't allow me to create any other AMI except Amazon Linux AMI.So here is the code that I am using to create an EKS, and one more thing is that I cannot get public IPs of the ec2 to connect via ssh. Thank you.

https://redd.it/oe3ul7
@r_devops

Are you in Devops? Can you share your opinions on challenges you are facing currently?

Hi all,

I am working on creating a software in the Devops Sphere. I have my own assumptions based on the problems I see in the industry but I want to avoid creating a software that nobody wants, therefore I would love to have your opinions and feedback on the current problems you are facing and also have the opportunity to validate my own assumptions.

Here are the main questions:
- What are the main challenges you are facing to implement a full CI/CD pipeline?
- What are the main challenges you see when deploying IaC?
- What are the main challenges to implement security in your SDLC?
- Where do you want to see a change in the Devops/Devsecops industry?
- What are the problems you wish to see fixed asap?
- Are they any other questions I should have asked?


Thank you for your time!

https://redd.it/odvx96
@r_devops

Languages used for SRE or DevOps daily professional work.

Languages commonly used for automating build piplelines, and ensuring services are kept running (up) in your work place.

View Poll

https://redd.it/oep9lv
@r_devops

Anyone has a easy to remember shortcut for 12Factors?

Hello Community,

Just wanted to check if anyone has already devised a super easy way to remember the 12 factors of as suggested in 12factor.net.

Those seems a handful to recall without a good acronym or phrase.

Cheers

https://redd.it/oeq0a8
@r_devops

Tara Report - do you use it?

What are your thoughts on it. This is what I have so far:

1) https://www.portshift.io/blog/what-is-the-tara-report-and-why-should-devops-pay-attention-to-it/

2) https://www.mitre.org/sites/default/files/pdf/11\_4982.pdf

3) https://www.ciscotechblog.com/blog/what-is-the-tara-report/what-is-the-tara-report-1\_hu4d3eee6b3a92e7a5966b23adb743e7fa\_71006\_960x0\_resize\_box\_2.png

Anything else to add? Have you used it?

https://redd.it/oerl15
@r_devops

Invalid Parameter Combination when creating a new Elasticache Redis Instance using Cloudformation.

I am trying to create a Redis instance from scratch using elasticache service, I keep getting the following error:

>Cannot use the given parameters when creating new replication group in an existing global replication group. (Service: AmazonElastiCache; Status Code: 400; Error Code: InvalidParameterCombination;

​

I am new to using elasticache within cloudformation, I would appreciate any help to solve this issue or understand the valid parameter combination, and excuse me if there are any beginners mistakes.

I am using the following snippet:

cachesubnet:
Type: AWS::ElastiCache::SubnetGroup
Properties:
CacheSubnetGroupName: !Join "-" , ["rb", !Ref Environment, "redis-subnet-group"]
Description: subnet group for redis
SubnetIds:
- !Ref private1a
- !Ref private1b
Tags:
- Key: environment
Value: !Ref Environment
redis2:
Type: AWS::ElastiCache::ReplicationGroup
Properties:
AtRestEncryptionEnabled: True
AutomaticFailoverEnabled: True
AutoMinorVersionUpgrade: True
CacheNodeType: cache.m5.large
CacheParameterGroupName: default.redis5.0
CacheSubnetGroupName: !Ref cachesubnet
Engine: redis
EngineVersion: 5.0.6
NumNodeGroups: 1
GlobalReplicationGroupId: !Join "-" , ["rb-horizon", !Ref Environment]
MultiAZEnabled: True
NodeGroupConfiguration:
- PrimaryAvailabilityZone: us-east-1a
- ReplicaAvailabilityZones:
- us-east-1b
- ReplicaCount: 1
PreferredCacheClusterAZs:
- !Ref private1a
- !Ref private1b
PreferredMaintenanceWindow: mon:06:30-mon:07:30
ReplicationGroupDescription: Horizon for env
ReplicationGroupId: !Join "-" , ["rb-horizon", !Ref Environment, rgi]
SecurityGroupIds:
- !GetAtt mainSecGroup.GroupId
Tags:
- Key: environment
Value: !Ref Environment

https://redd.it/oeseey
@r_devops

How to allow external applications to write on a ceph rook volume used by pods?

Hello, I have a kubernetes cluster with a ceph rook cluster that works.

Now I need that external applications can write on the volume shared by pods, but I do not find how to setup a fixed mount point to be sure that a file placed there(e.g. in /shared/volume1) is uploaded in the volume.

How can I do? Thank you

https://redd.it/oeu8yg
@r_devops

Day to Day life - Public Sector DevOps

Day to day life - Public Sector DevOps

Hi All, long time reader first time poster. Seeing as there’s often lots of “how do I get into DevOps” type question on here I thought it would be useful to give an overview of a typical day of a DevOps engineer for a public sector company. I’ve tried to add a bit of humour in too.

0730 - You wake from your slumber. Having told yourself you’d get an early night you stayed up till 1am gaming. The light stings your eyes with the power of 1000 suns. You power on as it’s Friday and the last day of the week!

0800 - having showered and eaten you grab your bag and leave the house. You jump in your car. You begin to reverse out and then realise you’ve left your building pass inside and have to run in and get it. During your drive you try to remember if you pushed the changes for an ansible playbook a rather grumpy project manager requested yesterday to be ready for last week.

0830 - you arrive at the office and find there’s no space in the on site car park because for some reason people get in an hour early. You’re forced to the parking annexe which is a half mile walk away.

0845 - You arrive at your desk and stick the terrible laptop issued to you in its dock. Someone had adjusted your chair thinking they’re funny. You say morning to the rest of your team.

0900 - You open your emails for the first time today. You discover that no, you didn’t push the changes and now the project manager had escalated it to your boss.

0930 - Stand up. You enter the meeting room and chat with your colleagues. Your scrum master then turns up 10 mins late and says
“Before we start, we’ve been asked to write a Python script for key business project. They need this by the end of the day. Any volunteers?” The room goes silent. He gives you an awkward look “ed11e?” He’s just ruined your Friday in one blow. Gilbert and Sullivan’s sound of silence starts playing in your head. It’s always you.

1010 - Standup overran as usual and you didn’t listen to anyone’s updates because you’re still salty from being given the Python task. You speak with your scrum master and find the contact for the script.

1015 - You grab a coffee and return to your desk. You setup a call with the requester for the Python script. They decline the invite and propose a new time of 1500 on a Friday. “Can’t be that urgent then” you exclaim. Some of your colleagues laugh. Why do some people not understand productivity is at its lowest on a Friday afternoon. This has ruined your second Friday in a row after last weeks production incident on your source control system which kept you up till 11pm resolving and meant you didn’t go to the pub.

1030 - Your asked to complete a Pull request review for a colleagues Jenkinsfile. It’s really long so you sit with them to talk you through it. You give them some feedback and they are grateful. Your day is looking up!

1100 - The big monitoring board turns red with an alert. Jenkins has decided to fall over and now no one in the company can build or do any deployments. Your boss comes over and says that this is now the number 1 priority to fix as it’s blocking production deployments. Your team has requested people stop doing production deployments on Friday afternoons for months. No one listens.

1110 - Having confirmed no one can SSH into the server you raise an incident with the infrastructure team. They say the server is down because of some patching work no one told you about. You mention to them that your Jenkins system is listed as a production server and should be patched out of hours. They say “oh yes, it is, we’re sorry, it will be done in 2 hours”. You send an announcement out to users advising them and get lots of angry replies.

1200 - Deciding that no work can be done until Jenkins is fixed you decide to go to the canteen for lunch. You’re team is accosted by a product manager for eating when they “should be fixing Jenkins”. It takes all your strength not to punch him in the balls. You explain to him that it’s not your fault. He walks off

with a “it’s not good enough”. You mutter “jog on” under your breath and hope he doesn’t hear it.

1300 - You return to your desk. Jenkins is still dead. You phone the infrastructure team and ask for an update. They say the patching has been completed but the server won’t start. You escalate to the infrastructure manager.

1315 - Jenkins is alive again. The server couldn’t start because of a data centre issue.

1330 - The build queue on Jenkins is ridiculous. People start complaining they can’t deploy. You polite tell them to be patient.

1400 - A developer comes over and says the source control system isn’t working. It is, he just hasn’t pulled the latest changed in. You send them a link to the git documentation. They won’t read it.

1430 - A developer raises an incident saying Jenkins is broken as their build is failing. You look through the stack trace and see a test has failed. You send them a dump of the error. They ask why the test is failing. They don’t know. You tell them to talk to the rest of their team who wrote the tests as you don’t know how they work.

1445 - Another failed build incident raised. This time for a genuine incident. There’s a bug with the pipeline and a docker image needs a few tweaks. It’s a 10 min job and will save lots of pain but you can’t do it as your scrum master will get upset you’ve messed his JIRA reports up. You raise a ticket in the backlog and tell your scrum master it needs fixing urgently. He acknowledges it. This means it will be fixed in 2 weeks because he doesn’t want to fill the sprint up.

1500 - Python script call. You join the call and find out that the product manager has said it’s key to the business to get it through. It isn’t key to the business.

1530 - You throw the script together in such a way that would make the PEP8 maintainers cry. You don’t care. If they want it done quickly they can’t expect good practices and you just want rid of them.

1630 - The Python script is finished and you hand it over. They seem happy with it. You say you’ll tidy it up another time. You won’t.

1700 - a colleague suggests going for a beer after work. You contemplate resigning immediately so you can go for one, but you’re stuck sorting out the mess from the Jenkins incident earlier.

1800 - build queue cleared and all the incidents are closed. Result! You can meet your colleagues in the pub. You move your car from the annexe to the main car park as you’re intending to leave the car overnight so you can de-stress.

1830 - you arrive at the pub and the product manager from earlier is in there. He gives you a look of complete destain. You don’t care, you’re here to get drunk.

TLDR: standup, meetings, production incidents, lunch, late requests, users who don’t know what they’re doing.

https://redd.it/oevta7
@r_devops

Writing my report (College) and getting lost/confused between Devops is a methodology and Devops is a culture.

I am writing my report for Programming "POE" and, upon my research, i am getting conflicted between Devops is a methodology or Devops is a culture. Is it both, one or the other?

https://redd.it/oev4ob
@r_devops

Someone please save me from GCP's god awful billing

I've been an AWS fan for obvious reasons but their granular billing flies under the radar. You can see what service used how much and the relevant amount that you were charged for it.

Google on the other hand, wants me to export the damned billing data to BigQuery and analyze it there. Why?! Why can't I see how long I ran the type of VMs for? Or how much Cloud storage I'm using?

Why can't I just see what you've so graciously tracked and slapped onto the credit card?

https://redd.it/oexmis
@r_devops

ICYMI: Phabricator is no longer actively maintained

Phabricator is no longer actively maintained as of June 2021. The company that back this product (Phacility) is winding down operations.

https://redd.it/oexmvu
@r_devops

Rundeck on K8s - has the industry moved on?

My team is working on upgrading our Rundeck deployment on EKS. We have an older helm chart and were considering just moving to the latest maintained one, but when we went to search for one, we noticed that most helm charts for Rundeck were depricated around early 2020. Did Rundeck lose it's niche while we weren't looking? If so, what are people who operate in K8s and AWS environments tending to move toward? If not, can anyone point me at a well maintained Rundeck helm chart? Thanks!

https://redd.it/oexde3
@r_devops

Angular vs React: Which js framework to pick for your project in 2021!

https://arohi-adhyaru.medium.com/angular-vs-react-which-js-framework-to-pick-for-your-project-6f57ef8cda39

https://redd.it/of1wl5
@r_devops

Transitioning to DevOps - what do I need to know?

I'll cut to the chase. I've had a fantastic opportunity arise for an internal role change. Going from IT Guy to DevOps. My intended career path was to end up doing DevOps/ SRE in come capacity, so not completely blind to the field. Somewhat pooing myself though, as I have a elementary knowledge of Unix, and can only write basic scripts in Posh. Rest of my exp. is desktop supp. & MS365 w/ some server work.

Oh, and I need to be up to speed in about 3 months. Well, at least able to produce work of some value. I know I won't be an expert.

In terms of tools and languages, my main priorities are Unix command line, Python and GCP (what we use). I've had some exposure to GCP, so that's not so scary. The role is in the biotech sphere so also need to learn more about computational biology some time.

All very exciting and an opportunity I absolutely need to crush.

So, what's next?

I need to learn:

DevOps Methodology
Python - have a live Pluralsight membership so going to get started there.
Unix/ Bash - as above
GCP - Get stuck in with their CI/CD tools, adding Docker and Kubernetes under this learning umbrella

My questions:

Any recommended reading or learning sources you have found useful in your journey?
I plan to read the Phoenix Project but have also started reading the Linux SysAd Handbook
Any useful Python libraries you wish you had known about or find useful day-to-day?
Any encouraging stories from people who have made a similar jump (please god)?

Of course, I will be checking in with the team as to what they think I need to know and in what order as soon as I can, to avoid wasting any time.

https://redd.it/of11cw
@r_devops