Google Cloud Network tiers, Jenkins plugin for ACI and ACS, and the alliance between Alexa and Cortana.
Google Cloud Networking
Google introduced a second tier for their GCP network layer. Unlike what you’d usually expect however, this is not a premium offering, but a lower level. One of the strengths of GCP’s networking, which frankly I’ve overlooked many times myself, is that traffic goes through Google’s network for as long as possible. As shown in the above article, traffic enters Google’s network at the closest point to the user and then travels through that network to wherever your compute instances are. This is what allows them to offer certain features at a global level, like a global load balancer, that other cloud providers can’t.
Now however, they have introduced the ability to let traffic route through the public internet to get to the target region. This means a lower price, and is comparable to other cloud providers in what it does and how it works. I assume GCP introduced this as a cost saving measure for regional clients1, but I don’t think they’re actually happy with it. Reading the announcement I get the feeling that they believe this to be an inferior solution2 that should only be used if you’re regional and believe costs are the most important thing.
Jenkins for containers in Azure
Next up, I want to have a quick look at some Jenkins plugins that were released by Azure. The headline plugin is for enabling you to easily deploy to ACS regardless of which orchestrator you use3. This is nice, but to me the other plugin is far more interesting.
When Azure announced their Azure Container Instances, I really liked their potential and the new Agent Plugin for ACS and ACI shows some of that potential made real. In short, this allows you to use ACI4 to run your jobs on. As many others, I’ve long switched to running my various build jobs in containers but this plugin prevents you from needing to manage the underlying hardware. In effect, you can have a low-powered master Jenkins instance that spins up “serverless” Docker containers for your build jobs when needed. This solution is fast, cheap, and flexible. A perfect combination for when you’re not continuously running build jobs.
Cortana and Alexa
Microsoft’s Cortana and Amazon’s Alexa will soon be able to talk to each other. Being able to combine assistants and use each to their strengths sounds like a step in the right direction for better results, even if you’ll need to get used it.
Unfortunately it sounds like initially you’ll have to explicitly tell them to talk to each other. I think that having to say “Alexa, open Cortana” before the actual command doesn’t make sense from a usage perspective and will stop people from using it. Addressing Alexa on your Xbox, or Cortana in your Echo, directly however is a different matter. That sounds quite appealing and hopefully is where this will eventually lead.
- Many Australian traditional companies for example don’t have much of a presence outside of here, so they wouldn’t want to pay for global traffic. [return]
- Which then leads to other cloud providers being similarly inferior. [return]
- Unlike GCP or AWS, the Azure Container Service gives you the flexibility to use either Docker Swarm, Kubernetes, or DC/OS as orchestrator. [return]
- And ACS, but that’s less interesting in this context. [return]
Related content you might like:
- Week 4, 2018 - Go in AWS Lambda; AWS Auto Scaling; Cloud AutoML
- Week 31 2017 - CloudFormation StackSets; Microsoft Container Instances; Flash EOL
- Week 6, 2018 - Kubernetes in Docker; CoreOS Joins Red Hat; Nintendo Switch Online
- Week 5, 2018 - Meltdown and Spectre Updates; Amazon Go; Apple HomePod
- Week 49, 2017 - re:Invent 2017; macOS root bug