Capacity enhancement is part of cluster sizing and maintenance. Proactive cluster sizing is always recommended. Commissioning a new Node or set of node are planned in advance to ensure cluster availability. It’s recommended to have some extra resources available. Majority of the time job troubleshooting is not in scope of admin team. Hot fix, trouble shoot, bug fix, change, enhancement, development need to be scoped carefully. Organizations new to hadoop are maturing in terms of scope of work, team and sizing etc.
Below is a 2 node HDP 2.5.0 cluster. It’s a demo cluster and in production numbers looks different. Screen shots of 2/2 data nodes live. HA, QJM, Recovery, security, access not yet implemented in this demo cluster. Don’t get me wrong, different issues one can face while commissioning a new node is challenging. Bigger cluster pics are on way, soon i will share it across as and when possible.
Below is a 3 node HDP 2.5.0 cluster. It was resized from 2 nodes to 3 node cluster. Screenshot showing 3/3 data node lives with other metrics. It’s a demo cluster and in production it looks different.
Legal: Document cannot be reproduced or used without owners consent.