jrgnsd

SEEKING WORK | REMOTE

DevOps specialist for hire

Experienced in creating and automating AWS infrastructure

Let me help you to * Create your cloud infrastructure from scratch, or formalize your current infrastructure * Advise on the best / cheapest / fit for purpose setup for your app or website * Automate pipelines and deployments * Automate scaling, monitor and alerting * Integrate with communication and management systems * Reduce notification fatigue and increase effectiveness of your alerting

https://jrgns.net

Ruby gems (libraries) aren't pulled feom github. Yes, you CAN pull it from there, but most public and published gems don't come from there.

SEEKING WORK | from Johannesburg, South Africa | Remote | up to 10 hours per week | $60-$80 per hour

Development and Data specialist looking for problems to solve.

Data Problems: Anything involving Elasticsearch or data migrations, scraping and cleanups.

Development Problems: Ruby, Sinatra, Sidekiq and even some PHP.

See https://github.com/jrgns, https://github.com/EagerElk and https://jrgns.net

Contact: jrgns at jrgns dot net

Location: Johannesburg, South Africa

Remote: Yes

Willing to relocate: Possibly

Technologies: Ruby, Elasticsearch, Postgres, MSSQL, C#, Micro Services, Cloud Infrastructure, ...

Résumé/CV: https://bit.ly/3eRya5Q

Email: See CV

Github: https://github.com/jrgns

http://hackerpla.net/requestd/

RequestD - An Elasticsearch backed dashboard and interface to track and monitor websites and APIs. Great tool if you need to monitor error rates and other SLA related metrics for a web service.

An interview is more about how you present yourself than about your actual skills, so make sure that you create a good impression first before telling them. In my interviews I usually ask about a record towards the end of the interview so that I'm not biased, but if it comes up earlier, be truthful about it.

I've done a couple of contracts through them. For the most part it's beem a good experience.

They try to vet the clients as well as the coders / talent, but it's not perfect process so sometimes a dud falls through the cracks. They do have solid procedures for both the client and freelancer to follow if either party isn't happy, which includes the person who set up the engagement initially.

Much better than the other sites.

Project: I'm building a simple management interface on top of Elasticsearch. I find their pricing model where you can only get some key enterprise features (like security) when you take a support package somewhat restrictive. It's still early days, but I'm getting there :)

https://github.com/eagerelk/proxes

My experience working through Toptal has been mostly positive (although I haven't done a contract through them in quite a while).

I failed my initial application due to my inability to solve a couple of algorithm specific tests as other commenters have mentioned. I reappllied a couple of months later, did a couple of codility tests to ensure I'd pass this time, and flew through the rest of the process.

Yes, their signup process will potentially disquallify a couple of good devs, but it will also weed out a lot of bad devs that can't take the time to learn the solutions to common algorithm problems.

Once in, they have a variety of jobs available, some good, some bad. Overall I don't mind working for them. I've been happy with the Upwork dev's I've hired so far, so can't comment on if I'll hire through Toptal.

Thanks for linking to the article. I remember reading something like that but had no idea on how to find it.

Setting up nginx to proxy to Elasticsearch with HTTP auth on top is fairly trivial. There's a couple of good articles on the web if you google for it. Also, should you have an Elasticsearch support contract, you get access to the Shield plugin which has extensive access control.

But yes, the fact that it is open OOTB is frustrating.

Although I like the idea behind this, I find it quite difficult to get started in using it. Any good posts or guides for that?

The title doesn't match that of the article, and sounds somewhat biased?

It looks like since you define both the queries and the inputs, you can set it up to support SNMP. You can use Logstash to listen for SNMP traps, and push it to your OpenTSDB or Elasticsearch instance, from where Bosun will pull it.

SEEKING WORK - REMOTE (Based in South Africa)

I wrangle data as a hobby. Lately I've been playing around with machine learning and how it can be applied using Elasticsearch.

Contact me if you need anything around the ELK stack - Elasticsearch, Logstash or Kibana.

Contact details on my profile.

Github: https://github.com/jrgns Blog: http://blog.eagerelk.com PoC: http://filter.eagerelk.com

You train your own set by posting the document and the classification / category to the service. The training set is then used to calculate the prior probabilities needed by Bayes theorem. Once you have a proper training set, you can post a document to the service and it will return a list of probable categories from the training set.

Your set might look something like this:

Category - Document

negative - I don't like ice cream

positive - That's an awesome idea

neutral - The wind is blowing today

positive - We won!

You'll get a categorization of either positive, negative or neutral.

I find that having a bayes classifier handy can help a lot with categorizing documents or text. I manually keep track of my bank accounts using a spreadsheet, and I started using a classifier to automatically categorize the transaction. It's quite a time saver.

I recently had the need for Bayes Classifier[1] in a couple of projects, so I wrote a service that exposes one through an API. You can set up your prior set and then get predictions against that set.

I haven't gone through the trouble of making it suitable for public consumption yet. Would anyone be interested in consuming such a service?

[1] https://en.wikipedia.org/wiki/Naive_Bayes_classifier

I have a couple of Zapier zaps set up to alert me of hot posts in a couple of niche subreddits. I use these alert mostly to inform my thinking and content for a blog I'm working on. It helps me decide what content will add value as I can see the type of questions people are asking, and it allows me to promote my blog as I can reference it in comments / answers.

Most of my interaction with Reddit is driven by these alerts.

I inherited a similiar system. It was painful to say the least. I didn't realise how painful until I started using proper queueing systems and Sidekiq.

I think you're referencing this article about MySQL job queues: https://blog.engineyard.com/2011/5-subtle-ways-youre-using-m...

There's a number of solutions, depending on your stack: at and cron are good OS level solutions. Sidekiq is great for Ruby.

I think the problem lies in that PHP wasn't designed to be a long running process, and for a scheduler you need something long running.

SEEKING WORK / South Africa / Remote

I provide solutions focused on the ELK stack: Elasticsearch, Logstash and Kibana. The main use cases are either full text search or time series data, such as logs and transactions, analysis and search.

If you're still searching through your logs and data using %TERM% or grep you need to speak to me.

More info: blog.eagerelk.com | info@eagerelk.com | @eagerelk

When I initially saw the ad, I thought it was satire. Didn't think it could be real.

The clean and crisp look of the site is stunning. The fact that there's no busy sidebar cramping up the content makes a huge difference.

Does anyone know if it's an available theme, or if it was custom built?

It's the first time I see github's Readme's being used as a blogging tool. Is this common? I've started to link to a Vagrant/Ansible repo for my setup / code intensive posts, but having the code and the text encapsulated as a repo is quite novel.

Got it working on ST 2 on Windows.

Ensure that you're editing the right flavor file, and restart ST for good measure.

I find it ironic that a guy with the surname vogel (the German and Dutch word for bird) writes about Flappy BIRD.

I've always loved grep. Now I know why.