Unless you have explicitly written the data the first time in the byte-representation which Phoenix requires, you cannot treat the data as any other type than VARCHAR. This is not simply a "best practice" -- this fundamentally does not work when you write the data in a different representation than Phoenix expects. ... View more

You're exactly right that you shouldn't delete things by hand 🙂 If you're on >=HDP-2.5.x, make sure to disable the HBase backup feature. This can hold on to archived WALs. You'd want to set hbase.backup.enable=false in hbase-site.xml. If you have HBase replication set up, that's also another potential candidate for why those files are not being automatically removed. Lots of HBase snapshots are another candidate (like Sergey suggested already) -- drop the old snapshots you don't need anymore). Turning on DEBUG in the HBase master should give you some insight to the various "Chores" that run inside the Master to automatically remove (or retain) data. ... View more

You would need to implement a custom Coprocessor (likely, a RegionObserver would be sufficient) to notify any external system of changes to HBase. This is tricky to do correctly, so I would recommend that you re-think your architecture and make sure this is the way you want to implement this. https://hbase.apache.org/book.html#_types_of_coprocessors (For example, you may find it easier to push all of your data through Kafka and instead send events to HBase and HDFS per your business rules) ... View more