kevinrr888 opened a new issue, #6030:
URL: https://github.com/apache/accumulo/issues/6030
**Describe the bug**
**Background:**
#5990 fixed a bug where users could use `NewTableConfiguration` to attach an
iterator or set a property that would conflict with table defaults (i.e., the
default VersioningIterator at priority 20 for all iterator scopes, and the
default table constraint (`table.constraint.1` as `DefaultKeySizeConstraint`)).
For example, previously a user could attach an iterator (or manually set a
property) through `NewTableConfiguration` to create another iterator at
priority 20 leading to a table with 2 iterators at priority 20.
Another example, a user could set `table.constraint.1` to `some_constraint`,
leading to a table without `DefaultKeySizeConstraint` or without
`some_constraint` (nondeterministic-which gets applied?)
**The bug/what needs more investigation**
There are other means of setting iterators. For example, setting a scan
iterator with `Scanner.addScanIterator` or having an iterator or property
inherited from the namespace level.
Need to investigate other means of setting iterators/props that would lead
to conflicts with default iterators/props and fix any issues found.
**Versions (OS, Maven, Java, and others, as appropriate):**
- Affected version(s) of this project: 2.1+
**To Reproduce**
For checking scanner:
Create a table, create a scanner for that table, try to add an iterator to
the scanner that would conflict with the default VersioningIterator, expect it
to fail
For checking namespace:
Create a namespace with a property or iterator that conflicts with table
defaults, create a table under that namespace, table creation should fail or
potentially creating the namespace should fail
**Expected behavior**
Two iterators with same name or priority should not exist simultaneously
Need to prevent nondeterministic results of setting props
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
