Hive Mind Your Frontend
Updated March 23, 2018

Hive Mind Your Frontend

Vlad VARNA | TrustRadius Reviewer
Score 10 out of 10
Vetted Review
Verified User

Overall Satisfaction with Docker

We are using Swarm for our analytics gathering service. Using swarm allows for quick workload scaling and using less hardware than was needed before.
  • Creating complex containers using docker files which automate a lot of DevOps manual labor
  • Having some preconfigured containers to do fast tests
  • The swarm takes away a lot of the work you would need to do for high availability
  • Kitematic UI is still very limited in functionality
  • Containers on Windows are somewhat hit and miss, Linux is strongly recommended
  • Swarm interface is mostly command line
  • Some network limitations (like remote client IP passthrough)
  • With Swarm we managed to have the horizontal network bandwidth scale to gather data from clients on slow internet connections
  • A minor negative is that you always need at least 3 nodes for redundancy, which makes it less ideal for quick proof of concept type scenarios.
For light weight frontends, the Docker Swarm is the easiest to manage. One person can then mange the production environment and at the same time develop new software.
Great for light frontends and (REST) microservices that don't depend on hardware/drivers and just do DB/file IO. Not so great for dev virtual machines and testing complex network configurations.

Using Docker

20 - Product Management, R&D team
  • lightweight scalable services that require a lot of concurrent access
  • load balancing
  • high availability
  • devops

Evaluating Docker and Competitors

Yes - HAProxy and VMware/Cirtrix VMs. Manual administration was too cumbersome. Resource utilisation was not optimal.
  • Product Features
  • Product Usability
It was quick to learn, and automated a lot of the manual devops work needed before
If I had more time, I would have tried Kubernetes as well.

Using Docker

Kitematic is still in alpha.
The CLI is somewhat unintuitive at first.
ProsCons
Like to use
Relatively simple
Easy to use
Technical support not required
Consistent
Quick to learn
Convenient
Feel confident using
Lots to learn
  • Scale the number of replicas
  • Template a service
  • Install the service
  • tagging the containers takes some getting use to
  • some networking details are hard to configure (like remote client IP passthrough)
  • some service manipulation commands have bad defaults (like changing an image but not using the authentication needed to pullit by default)