Follow

I finally got stable releases in the rpm/deb repository for BigSense.io. Stable release pushed to docker hub too. I've still got a ton of documentation to update, I need to migration the current VMs docker, and I haven't even looked at the lab manuals yet.

But I should probably break and work on my presentation for the Ruby meetup next week instead. That's going to come up fast.

@djsumdog
Ooh do you work with sensor networks?

@meff Yep, I maintain BigSense.io, an OSS tool used at the University of Dayton for temperature, humidity, GPS and rain water sensors.

@meff Right now just temperature average. It use to do averages for the water sensors but we pulled a lot of that out when the startup went under (the founder wanted it in and a lot of it was super specific use-cases).

@djsumdog
Ah gotcha. At least y'all open sourced the property of the startup

@meff Yep, that was one thing I made sure I could do; knew that startup wasn't going to last. The videos are a bit out of date. I've done more recent presentations where I showed how to clean up the Scala code to make it more functional, but I didn't get videos of it. The first 10~15 min shows the project we were working on though.

@djsumdog
Ah Scala with regression tests in python. Real cool!

@meff Yea once I get all the docs and dependencies up to date, I want to add ScalaTests in as well; possibly add support for the H2 database as well.

@djsumdog
We use H2 as a test db for our Scala stuff at work and it works fairly until it has differences with MySQL.

@meff Yea I ran into the same stuff when I worked at GettyImages in Seattle. At my last job we used docker containers to we could test against a real database; worked out a lot better .. didn't implement things twice that didn't need to be. 😋

@djsumdog
Hm did y'all run Scala in a container?

@meff Yea at GettyImages we did; Scala containers deployed via Marathon/DCOS. My last job was similar, but Java + K8s instead, and not as slick/flushed out.

@djsumdog
Huh interesting. I always thought running JVM stuff in a container was extraneous, but it probably makes local testing a bit easier

@meff It helps with consistency, immutability, rolling back/forward. OpenJDK containers are kinda big, but they're layered even in the docker repos, so the size isn't that bad. Debugging in a container is a bitch of course, in any environment (even with tools like sysdig). You have to be careful with the JRE and cgroup memory limits, but they're wrapper scripts for that. I wrote a whole post on what I think about containers: penguindreams.org/blog/my-love

Sign in to participate in the conversation
Mastodon

Hitchhiker.social is a Mastodon instance created by Sumit Khanna, a technologist and blogger out of Chicago. This is an experimental instance that's currently invite only.