vROPs and True Visibility Suite for APM

Hi all! Long time listener, first time caller.
Looking for some insight from this community to see if there is any familiarity with using vROPs and True Visibility Suite for their application monitoring.


Some background:

At the moment my company is using Dynatrace and is using it quite poorly. We have run into a scenario that the DevOps teams are too busy to learn the tool proprerly and due to it's high cost we are really not seeing the ROI. This has really put on the pressure to either find a way to upskill or retire.

We are also mid conversations with Splunk for logging as we currently have no logging solution outside our external hosted SIEM. They have been trying to sell us on their Observability offering but considering teams don't configure something like Dynatrace properly, the likelihood of them properly instrumenting that part of Splunk seems like a stretch.

The question at hand:

vROPs for infrastructure monitoring is really quite cheap and our VMWare SME swears that it is one of the best tools for anything VMWare. I can see that the cloud edition (which we are moving to) has a number of content packs for APM and there is also the ability to purchase additional packs through True Visibility.

I realise that RUM and Synthetics will not be covered by this solution, but wondering if anyone has any experience with using these products for APM? Does any one have any issues or thoughts on the products?

Also happy to hear other suggestions.

https://redd.it/p97dpz
@r_devops

where to place deployment configuration

Hey everyone, I am currently working on building a CI/CD pipeline for my team who currently is used to deploying applications to production from their local machines.

I have been able to demonstrate how a CI/CD pipeline would work and have gotten buy in from the rest of the team. One of the issues which we are facing is where to store the deployment configuration, for example: the AWS Region, OS limits (CPU / RAM), number of replicas, etc. We currently have 3 environments (DEV / TEST / PROD) and we aim to size our applications differently on DEV than we do on PROD.

I see two options:

1. Have the deployment configuration be stored within the application's source code repository; or
2. Have the deployment configuration be stored within an external repository, and then checkout the repo at some point during the CI/CD pipeline.

The first option feels wrong to me simply because of the way Git works. For example, if my deployment configuration for DEV has to change to downscale the machine's OS limits, that commit would be included as part of the next application's release when branches are merged from "develop" to "master". One benefit of this approach is that it keeps everything "close together", and any changes can be kept within the same repository.

The second option feels like the right approach because it separates application from deployment. I feel that the two have independent lifecycles but oftentimes impact each other. For example, an application may be deployed multiple times to the same environment but only built once. On the other hand, a change to the application might require a change to its deployment (such as increasing of its OS limits due to a more complex ETL job). Similarly, if a sidecar needs to run alongside the application to target some non-functional requirement, this can be done independently of the application code. In this case, storing the deployment configuration elsewhere from the application code repository makes sense. One benefit of this approach is that the "release" part may be managed by a different team who is authorized to carry out production deployments, however, if the release team is the development team, this might just slow us down unnecessarily.

I'd be happy to hear your ideas on how best to approach this scenario. Thanks for your time, looking forward to your replies!

https://redd.it/p9891d
@r_devops

How to Become a Certified Kubernetes Application Developer

This guide is a summary of my study notes for the Certified Kubernetes Application Developer (CKAD) exam that I recently passed.

Even if you’re not interested in the certification, consider this as your one-stop shop for Kubernetes: you have all the main technical concepts explained along with a myriad of examples in one place.

Furthermore, it has some extras that come from my experience preparing for and taking the exam.

https://redd.it/p9bgsx
@r_devops

Connect AD to AppStream 2.0 fleet

Just started a project and I was wondering where some of you would start if you were given this story.

AD and AppStream Fleets were previously created.

https://redd.it/p9e76i
@r_devops

Alternative to Rundeck / Selfservice GUI for non-devs

Hi community.

We've been using Rundeck internally to create simple "jobs" that enable non-devs (or non authorized personal) to run simple scripts, without them having database or other type of credentials.

Basically we are using Rundeck as a cheap/quick to deploy backoffice where QA, PMs, Support team can run jobs that query the database, or call APIs. This is working and it's simple to create jobs with inputs (text, date, dropdown).

Right now I'm looking for alternatives to Rundeck, not so much for it's automation or devops capabilities, but to achieve this goal: a web interface, with self-service jobs for less technical/dev teams.

Do you use Rundeck for this? Do you know any better/more friendly alternative?

I think there is a big market for this. Maybe some of the solutions of low-code/no-code on the market could also achieve this goal.

Thank you.

https://redd.it/p9f2r9
@r_devops

Free Docker registries

Since Docker announced that it would remove unused images from Docker Hub, I’ve been interested in listing places where I could host mines. As it’s for personal purposes, I’m interested in free plans. Here’s what I found.

https://redd.it/p9fi7w
@r_devops

Deploying a WordPress Site to the AWS Cloud like a Pro

https://towardsaws.com/deploying-a-wordpress-site-to-the-aws-cloud-like-a-pro-our-guide-for-painless-maintenance-using-5a3eff227095

https://redd.it/p9h84o
@r_devops

Create containers on events

My entreprise is using elasticsearch to store user data. I need to run containers on the events that exists in the elasticsearch database. New events are created Everytime. The containers will process the user data and add new fields on the event from elasticsearch that is being processed. New events can also be created from the services run in the containers if needed.

There are million events to process : users web accesses, users purchases, etc

The tasks run by the containers are very short : they take from 10 secs to 1 min to process.

The process would be the following (I think) : a deamon would be picking events to process and create pods on nodes which have a few pods running.

I want to be able to limit the number of containers working on specific events. For example : x containers working on users purchases, and x containers working on users web accesses.

As I saw, k8s is made to process normal services like a web server, not quick tasks as I want to do.

I saw that there are many tools that could be used for what I want to do like kubeless, knative, brigade, etc but the features looks almost the same and I don't really know what to choose...

Any advice ?
Thanks

https://redd.it/p9bjs8
@r_devops

Whats are the more common ci/cd pipeline builds

Hi, as the title suggests, I'm just after some insight from those who are in roles which include building ci/cd pipelines..

I just wanted to know;

- what sort of projects crop up more times depending/regardless of industry..

- the tools which are highly required ( terraform etc )


Thank you !

https://redd.it/p9kxwd
@r_devops

DataDog or AWS based monitoring solution?

So I am looking to build a monitoring solution for several AWS resources (redshift, EMR, EC2, etc.) and have the option to use DataDog. However, I have seen some suggestions to use AWS services to build such a solution. My understanding is cloudwatch is AWS's monitoring solution, but better dashboarding can be offered by Quicksight, so I was considering trying to pair the two services to have a cloud native solution, or a similar combination.

​

Has anyone seen anything similar to this? Or am I making this unnecessarily complicated by not just relying on DataDog?

https://redd.it/p9n7lh
@r_devops

Getting a DevOps Engineer job in the EU or UK as a US citizen?

Does anyone have any experience with this? How difficult is it to land a job and relocate for a couple of years?

The market is pretty hot right now and some of the salaries Im seeing here in NYC are crazy. However, there is a part of me that wants to see the world and explore but also continue to work as a DevOps engineer since I enjoy it. While the money sure is great here, I do feel like there is more to life. For reference, the places I was interested in are Madrid, Barcelona, and maybe London. How important is the language for landing a job? Id love to practice and learn Spanish fluently but don't think I could pick up that German easily.

As far as Ive researched you need a work visa to work anywhere in the EU (and the UK). Is it always done through the company? Im assuming here one would have to REALLY impress the company to have them go through the process of filing for visa sponsorship, right? Do companies even bother interviewing candidates from the US or just non-EU citizens in general?

In terms of salaries, I do see they are very low compared to US (using NYC,SF,Seattle as a reference for myself here). Is it possible to negotiate for more or can I generally expect under €100k/yr net (or even gross) no matter where I go? What are some of the tech hubs in Europe (I only know of London)? London salaries threw me off the most because afaik the cost of living is pretty much equal to NYC yet the salaries are almost halved in the UK.

Finally, for anyone who successfully relocated to the EU or UK, what is the process like in general?

Thanks!

https://redd.it/p9qrhv
@r_devops

Story defeated by Notepad++ (baby's first play book)

Howdy,

Last week was my first week on the job. I accepted a sysadmin role for an app team. The job uses devops methods to help bring different aspects of the this company's IT departments. I have done systems work and IT support before but this is my first big boy job. So while I am waiting on my company equipment to come in my manager gave me a pretty simple task. They basically just wanted me to create a simple Ansible playbook that would automatically update Notepad++ because he was tired of seeing the update pop up on his servers. No problem, it's a simple task because he understands I am fresh when it comes to Ansible so this is a good none destructive way to dip a toe. I hyped it up and was like no problem. I was like, "I am gonna get this done quickly and efficiently to impress my new boss. They will be like "Woah!""

Well life happens and for some personal reasons I am super stressed and to make matters worse it doesn't look like notepad++ has any hooks to just automatically update the app without having to have some manual intervention. (At least that I can find) so I am trying to juggle this very easy task and my personal issues and started to get frusted because the app has a third party update tool just not away to pass it commands.(again, at least that I can find) So I come up with the strategy to just simply do the task at hand and make a script that would Uninstall the application and install it using the new version from a share. Then rename the auto updater folder so the notification doesn't appear. The assignment was just so I could learn how to use the different systems anyway so I thought it was good to go. But now comes the fun, my playbook sucked! Since I couldn't use my stuff and was relying on slow VDI my code took forever to just type out so I resorted to just typing it on my machine and passing the code to remote app. Cool, my spacing was jacked and intentions off so now I have a ton of commits that are just like "debug...debug..debug." and a bunch of failed Ansible task that all the departments can see on a simple update. I finally got it to work and we are all good. I explained the situation and played it off like having automatic updates wasn't a good idea anyway as we should be doing something like "minus one version" to ensure security and reliability. (Which is the truth) I was just upset that something so simple turned out to be so frustrating, I am sure my personal stuff had something to do with it too as the script is super simple. But that is it that...that's baby's first playbook. I am now a Sublime fan.

TlDR: Baby's first Ansible playbook was more of a headache than I would like to admit. And I didn't get to look like a smooth hackerman at my new big boy job :(

https://redd.it/p9s41r
@r_devops

What are the various config files you need to watch for in a nginx-apache stack?

We use nginx for SSL, varnish for caching and then Apache for the php, so nginx to varnish to Apache and then finally php, and I am wondering what are all the config files to see how the ports and routing is done so that requests get redirected to the php.

https://redd.it/p9rcww
@r_devops

DevOps, AWS Contract Assignments in UK

I have more than 23 years of experience in IT and work as a Solutions architect. I am looking to start as a contractor in DevOps and AWS in UK. Do you think having such long running experience would help or hurt as a contractor? Are there any umbrella or consulting firms that I can check to start my phase of career? I have pretty solid technical and hands-on background.

https://redd.it/p9u93p
@r_devops

Thoughts on intentionally setting fire to raise fire safety awareness as nobody seems to care about the fire safety measures

I'm tired of advocating for best practices, availability, architecture document, etc., One queueing system is running in a single instance for 2 years and 20 production applications are using it. Highlighted it umpteen times that it's a single point of failure and business could be massively affected anytime, we should move to HA, yet nobody cares, nobody wants downtime.

Though it is recorded in the risk registry, I don't want to wake up in the middle of the night and asked to battle the fire. To follow the Dwight Schrute's footsteps, the thoughts of setting fire intentionally to see how far the impact is, how everyone reacts, often crossing my mind. This is not to demonstrate I told you so, but eventually the rescue operation will fall on my head when it goes down and I want them also to take responsibility.

What's your thoughts on this? Am I taking it too personal (sometimes yes)? Whats the best way to handle these kind of situations?

https://redd.it/p9vfy5
@r_devops

Sentry.io got 100k more transactions in one day and is maxing out my plan

Hello,

All of a sudden sentry.io started getting a ton more transactions. Seems I went from getting about 15 (20k per day) transactions per minute to getting over 80 (120k a day).

It seems sentry logs every http request that comes into my backend service. The single most common requests (by mode) are legitimate ones to valid routes on my endpoint, but the vast majority of requests are things looking for exploits (/phpMyAdmin). Those requests obviously fail (98%+ failure rate due).


I have just provisioned a AWS WAF for blocking obvious stuff (still getting /phpMyAdmin), but my services do need to accept GET and POST requests from third parties... can I have my WAF only whitelist certain routes?


Should I not have sentry configured for logging every rest request that comes into my backend?


Thanks

https://redd.it/p9vx4m
@r_devops

Some concept I try to understand with example

Hey,

So, we have a log server(hosted in ec2) written in Python which it's purpose it to write logs to some .csv file.

And it's only purpose is to see the logs through https://blabla/blabla.csv .

It is currently accisible through http (some solution the dev used for it), however very often we experience down time issues with it, so we starter looking for more stable solutions.

​

I've configured nginx server which all it does is : I've Created a symbolic link on the /var/www/html folder of that .csv file .

After testing it through the nginx, I retrive the file much faster (load time).

My question is -Does it what they mean when they ask to put it behin nginx?Should I move the entire app behind it? redirect the python code to the nginx?Or the "solution" I did can work.

Thanks for anyone willing to explain ::)

https://redd.it/p9wad1
@r_devops

Projects for a portfolio?

Hello guys, I am a DevOps engineer for the past 4 years working in a government agency meaning I can't take out any project I have made. What are some cool ideas for DevOps projects I should try? Thanks!

https://redd.it/p9xvfd
@r_devops

Solving the information gap between devs at the build stage and DevOps at the deployment stage

It's a big issue. I like the way this report solves the issue, but there are still some holes that need to be plugged. Thoughts?

https://ciscotechblog.com/blog/automated-policy-for-developers-using-cicd/

https://redd.it/p9ygpb
@r_devops

How to move from Manual to DevOps

Hello,
I need your advice on transitioning my career to DevOps from a developer role. Currently, I work at a small-medium company doing their operations like deploying manually to servers and sometimes doing developments with C#, etc. But my role is designated as a junior architect as I take care of mainly Website Deployments, database deployments, etc to different environments. By the way we are not cloud yet. I also take care of the TFS. Code reviews and check in before the deployments.
PS: When I say manual deployments, it is as building the Website/Api on the the Visual Studio and copy and paste to the IIS Website folder. We are a Microsoft shop.

Please advise me some steps to begin with, like I mean from scratch. Thanks.

https://redd.it/p9z6hm
@r_devops

Delivering B2B software as SaaS on Kubernetes

This post highlights the three key aspects that need to be considered when delivering a B2B software as-a-service on Kubernetes.
https://cloudark.medium.com/delivering-a-b2b-software-as-saas-on-kubernetes-2dbf8b88a7f5

https://redd.it/p9xxoz
@r_devops