#random (2020-04)

200iq , what a company

neat!

yet one more tool for @Zachary Loeber to add to his mile-long tool belt

I have been called a toolbag before…

accurate

[this] @Erik Osterman (Cloud Posse)

Hahah

haha! yes, will add the icons today (forgot!)

man I’d have to have a lot more confidence in a platform to spend $5k on a one shot than they give with a url link https://app.hunterz.io/srartupSignUp

Lets say we have a timeframe XXXXXYZXXX where 

• X is each 15 minutes

• Y is a deployment of a service

• Z is an event that needs to be investigated  I wanna know what did exactly happen right before Z happened I would then be able to see y happened x minutes earlier. The idea is to get all operational insights in one platform. Currently a lot of tools operate on their own and you might even be able to build something like this with kibana or grafana but the idea would really be to integrate all/a lot of operational tools to gain as much insights as possible.

For me it just becomes hard to open up all tools. For example database metrics I have in AWS - Cloudwatch (incl. alerts) but application metrics I have in prometheus inside each stage/environment and federated to one big monitoring cluster. It just becomes hard to use. But maybe I have approached these topics wrong and I just over complicated my setup but I have seen these setups across multiple companies.

The Idea would be if I get application alerts I can go to the timeline and see oh the database is currently having Issues so I maybe can start investigating there.

And if I would be the person responding to the Database high load alert and I can see oh that one service currently has a high load alert and I can start to investigate there. Because these are likely to be related.

Sounds like you want a single pain of glass like datadog or something where cloudwatch, datadog, timeline annotations from builds and all are displayed in single place.

Correct. I do think something like this exists.

I also would like add information to deployments like for example a package.json/lock to make it easier to know what really runs inside of my env

Look at datadog or something like that. I have used grafana and done loads of metric stuff . Don’t roll your own. Most of these services support annotations for events and tie build /release info into it. APM is expensive but some also track the assembly versions changing to automatically on apps allowing even more traceability on changes.

Id suggest something like datadog. If you prefer cloudwatch then you can also add cloudwatch annonations to metric timelines but I’m not a big fan of cloudwatch.

Also for what runs inside your environment, that is really separate metadata from monitoring related metrics

With AWS for example they have something called systems Manager. it can scrape inventory and give you all of the installed applications, their version number, drivers, services and more. That’s the service that’s designed to aggregate all of your information on what’s on a machine

And then there’s compliance manager which is when you set a required value or tool as needing to be installed to meet compliance and it can pass or fail. Lots of crossover and some of these things

Systems manager looks interessing. I for sure would not like to run my own metrics service more like an events aggregator is what I am looking for .. Which gives me a timeline from let’s say: Deployments, Bugsnag events, Alertmanager alerts, cloudwatch alerts etc.

My guess start with datadog. You need integrations galore. I plan on testing it out soon too. I’m concerned about how flexible it is with custom metrics but won’t know until I start too

Closing out stories

it does make it so easy to organize big community slacks..

Ever use terragrunt with the Azure provider?

What’s the difference between terragrunt usage for aws and azure?

well with terragrunt you still use terraform, but I support the use of more memes :)

https://i.imgflip.com/3y2ij7.jpg

Conceptually there is no difference but it was clearly written for AWS and there are very few terragrunt examples for Azure (or any other cloud for that matter). I’m also a bit leary of terraform wrappers that change the fundamental logic of how terraform works

@maarten in case of we are talking about binaries the meme is correct

What about Terragrunt do you like so much? More curious for our product team feedback than anything.

Jake… i want to run the same plan across multiple aws accounts without the need to reconfigure individual workspaces by cli for each. Haven’t seen anything that makes this seamless. I have to do some work around with inherited yaml files to try and make it more seamless

I can go into more detail but i’ve posted a lot about some of the challenges I’ve personally faced in trying to reduce code to make things more clean

I converted several dozen pipelines at my last project to this. Thank goodness it actually hit a general release or I definitely would have egg on my face…