Istvan Toth created PHOENIX-7317:
------------------------------------
Summary: Avoid re-using encoded column qualifiers
Key: PHOENIX-7317
URL: https://issues.apache.org/jira/browse/PHOENIX-7317
Project: Phoenix
Issue Type: Bug
Reporter: Istvan Toth
We have encountered a situation where the last column was dropped from a table,
and then another column of a different type was added.
This has resulted in errors as the new column used the same encoded qualifier
as the old one, and Phoenix tried to interpret the existing data as the wrong
type.
Perhaps the easiest way to avoid this would be to maintain the highest encoded
CQ ever present in SYSCAT and always assign a new CQ to newly added columns.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
