#44 https://blog.philipphauer.de/discussing-docker-pros-and-cons/
Disadvantages and Challenges
After praising Docker (too much?), let’s consider the drawbacks and challenges when using Docker:
Increased complexity due to an additional layer. This affects not only the deployment but also the development and build.
In addition, managing a huge amount of containers is challenging – especially when it comes to clustering containers. Tools like Google Kubernetes and Apache Mesos can help here.
The containers share the same kernel and are therefore less isolated than real VMs. A bug in the kernel affects every container.
Docker bases on Linux Containers (LXU), which is a Linux technology. Therefore, we can’t run Docker on other systems and our container is always a Linux system. But Boot2Docker enables the usage of Docker on Windows and Mac OS X by using VirtualBox. The Docker client runs on the host OS and communicates with the Docker daemon inside the VirtualBox. Unfortunately, this is less comfortable and makes daily use clumsy and more complicated than running Docker natively.
Introducing Docker can be a demanding and time-consuming task. We have to evaluate if it is worth the effort and the increased complexity. In a past project of mine, the management rejected Docker with the argument, that 3 nodes (production, pre-production and experimental) are not enough to justify the effort… However, when you have a cluster of nodes (when scaling horizontally) or use Continuous Delivery you need an approach to reproduce environments and Docker is brilliant in this. Moreover, Continuous Delivery reduces the time-to-market (new features can be brought to production faster because the deployment and environment setup is automated), which can be used to convince the management.
I also made the experience that there are also reservations about Docker in strictly regulated domains (like the banking sector):
To run a container you need root rights. This can be a problem for some companies, in which the colleagues that are supposed to run and update the application on the production server must not have root rights.
It is unclear if the usage of Docker agrees with certain security standards which have to be fulfilled (like PCI).
In some companies, there is the questionable rule, that only software from official/trusted sources can be installed on the machines. For instance, Docker is not included in Red Hat Enterprise Linux 6 and therefore needs to be installed from docker.com, which is an “untrusted source”.
https://www.quora.com/Is-Docker-ready-for-production
Further, from an ecosystem standpoint there are critical elements still missing or very immature:
Monitoring - This is something of a Docker issue as well as a full spectrum of metrics is only now coming on line for the currently line of tools to incorporate. (e.g. New Relic, Zabbix, Data Dog)
Logging - Companies that have sophisticated log strategies around customer usage behaviors and patterns find the STDOUT and STDERR paradigm to be far to limiting. Again the very recent addition of facilities such as syslog mean that the ecosystem will now be able to develop this sophistication (e.g. Splunk, Sumologic, Logstash)
#43 Iba a editar mi comentario anterior pero ya que estoy te contesto.
Aquí va un ejemplo real de uso de docker en producción, uno de tantos.
https://schd.ws/hosted_files/cloudnativeeu2017/e2/Kubernetes-From-Dev-To-Prod-GoEuro.pdf
Ahí desmonta las gilipolleces de compliance, escalabilidad, y ahora que lo mencionas motorización (porque por supuesto no existen los sistemas de motorización).