Today we continue discussing the best practice from storage engineering:
661) The operator used to build and deploy the storage server can take care of all the human oriented administrative tasks such as upgrade, scaling, and backups.
662) The logic of these administrative tasks remains the same across versions, size and nature.
663) The parameters for the task are best described in the declarative syntax associated with the so-called custom resource required for these deployment operators
664) The number of tasks or their type depend from application to application and can became sophisticated and customized. There is no restriction to this kind of automation
665) The containerized image built has to be registered in the hub so that it is made available everywhere.
666) There are many features on the container framework that the storage product can leverage. Some of these features are available via the SDK. However, container technologies continue to evolve in terms of following the Platform as a service layer and the public cloud example.
667) The type of features from the operator sdk used by the product depend on the automation within the container-specific operator specified by the product. If the product wants to restrict its usage of the container, it can take advantage of just the minimum.
668) One of these features is metering and it works the same way as in the public cloud. Containers are still catching up with the public cloud on this feature.
669) The operators can be written in a variety of languages depending on the sdk however in many cases a barebone application without heavy interpreters or compilers is preferred. Go language is used for these purposes and particularly in devOps.
670) There is no special requirement for performance or security from the containerization framework than what the application wants from the host because this is internal and not visible to the user.
661) The operator used to build and deploy the storage server can take care of all the human oriented administrative tasks such as upgrade, scaling, and backups.
662) The logic of these administrative tasks remains the same across versions, size and nature.
663) The parameters for the task are best described in the declarative syntax associated with the so-called custom resource required for these deployment operators
664) The number of tasks or their type depend from application to application and can became sophisticated and customized. There is no restriction to this kind of automation
665) The containerized image built has to be registered in the hub so that it is made available everywhere.
666) There are many features on the container framework that the storage product can leverage. Some of these features are available via the SDK. However, container technologies continue to evolve in terms of following the Platform as a service layer and the public cloud example.
667) The type of features from the operator sdk used by the product depend on the automation within the container-specific operator specified by the product. If the product wants to restrict its usage of the container, it can take advantage of just the minimum.
668) One of these features is metering and it works the same way as in the public cloud. Containers are still catching up with the public cloud on this feature.
669) The operators can be written in a variety of languages depending on the sdk however in many cases a barebone application without heavy interpreters or compilers is preferred. Go language is used for these purposes and particularly in devOps.
670) There is no special requirement for performance or security from the containerization framework than what the application wants from the host because this is internal and not visible to the user.
No comments:
Post a Comment