[
https://issues.apache.org/jira/browse/HBASE-27827?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17719831#comment-17719831
]
Niels Basjes commented on HBASE-27827:
--------------------------------------
I'm very happy you are picking this up.
Some projects at $dayjob now use BigTable which works great but is also seen as
too expensive for smaller projects.
Would be awesome if for those small projects we could deploy HBase instead
which would allow these projects to do it right the first time and seamlessly
grow from there.
So I have a few points that are on my "_Would be awesome if it could also be_"
list:
* Just yaml files
** So we can use this in cases where installing a new operator is not an option.
* Run it on kubernetes volumes (instead of HDFS)
* The option for running a single pod (non-HA) HBase
** perhaps even a single container?
* The option for running the smallest possible HA setup (i.e. 3 pods with basic
anti affinity)
> Introduce kubernetes deployment
> -------------------------------
>
> Key: HBASE-27827
> URL: https://issues.apache.org/jira/browse/HBASE-27827
> Project: HBase
> Issue Type: New Feature
> Reporter: Nick Dimiduk
> Assignee: Nick Dimiduk
> Priority: Major
>
> As per the
> [discussion|https://lists.apache.org/thread/fgxyk4y32xnhzr5prdmhfjkfpk15g5jx]
> on the dev list, introduce a basic harness for deploying ZooKeeper, HDFS, and
> HBase on Kubernetes.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
