[
https://issues.apache.org/jira/browse/HBASE-19775?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16324168#comment-16324168
]
Josh Elser commented on HBASE-19775:
------------------------------------
rubocop/rubylint stuff looks benign enough for us.
bq. As for the UT - honestly speaking I'm not sure how to check that
upper-level shell commands handled exception correctly.
I think that's fine. It looks like this happens when loading the next value
from ResultScanner errors out. I can't think of a good way to trigger this
either.
Looks like this UncheckedIOException came in via HBASE-17141 which is 2.0 only.
+1 from me. [~stack], may I land this on branch-2?
> hbase shell doesn't handle the exceptions that are wrapped in
> java.io.UncheckedIOException
> ------------------------------------------------------------------------------------------
>
> Key: HBASE-19775
> URL: https://issues.apache.org/jira/browse/HBASE-19775
> Project: HBase
> Issue Type: Bug
> Components: shell
> Affects Versions: 2.0.0-beta-1
> Reporter: Sergey Soldatov
> Assignee: Sergey Soldatov
> Fix For: 2.0.0-beta-2
>
> Attachments: HBASE-19775-2-branch-2.patch, HBASE-19775-branch-2.patch
>
>
> HBase shell doesn't have a notion of UncheckedIOException, so it may not
> handle it correctly. For an example, if we scan not existing table the error
> look weird:
> {noformat}
> hbase(main):001:0> scan 'a'
> ROW
> COLUMN+CELL
> ERROR: a
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
