Page MenuHomePhabricator

Draft a project plan for the Hadoop version 3 upgrade
Closed, ResolvedPublic

Description

We know that we need to upgrade our Hadoop clusters from version 2.10.2 to a more recent version, probably 3.3.6 or 3.4.0

In addition, we need to make sure that all of the services that are integrated with Hadoop either continue to function, or can be deprecated.

These services include:

  • Hive metastore
  • Hive query engine (potential to deprecate?)
  • Spark
  • Flink
  • Sqoop (Does not support Hadoop 3, but can allegedly be made to work with wire compatibility to Hadoop 2)
  • Gobblin
  • Presto
  • Druid

The purpose of this task is to track the initial creation of an upgrade plan, upon which we can then iterate.

We also need to make sure that we have sufficient safeguards in place in order to avoid data loss on HDFS.

Event Timeline

This task may be a useful reference. It was an investigation of what we considered to be unrecoverable and requiring a backup, before the last major operation to HDFS. We ended up with a figure of ~ 400 TB at the time.
T260409: Establish what data must be backed up before the HDFS upgrade

BTullis raised the priority of this task from Medium to High.Nov 26 2024, 11:11 AM

I have finished a first draft of: Hadoop 3 upgrade: high-level plan.
This contains a technical outline of the plan only, from which we can scope out phabricator tickets if and when we decide to proceed.

There is also a Decision Brief - Hadoop Upgrade document, which sets out the reasoning and other aspects that go into making a go/no-go decision.

I'll resolve this now, since the draft plan has been written and has been shared for review.