[
https://issues.apache.org/jira/browse/HBASE-16774?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15555823#comment-15555823
]
Jonathan Hsieh commented on HBASE-16774:
----------------------------------------
A few nits in the ruby parts:
{code}
54 @Test
55 public void testRunNoClusterShellTests() throws IOException {
56 // Start ruby tests without cluster
57 // System.setProperty("shell.test",
"test_connection_no_cluster.rb"); // REMOVE THIS LINE?
58 jruby.runScriptlet(PathType.ABSOLUTE,
"src/test/ruby/no_cluster_tests_runner.rb");
59 }
{code}
Add comment near top of no_cluster_tests_runner.rb saying intent of file (e.g.
for testing shell with hbase is not up, pulling in all *no_cluster tests from
dir xxx)
otherwise ruby parts look good to me.
> [shell] Add coverage to TestShell when ZooKeeper is not reachable
> -----------------------------------------------------------------
>
> Key: HBASE-16774
> URL: https://issues.apache.org/jira/browse/HBASE-16774
> Project: HBase
> Issue Type: Improvement
> Components: shell, test
> Reporter: Esteban Gutierrez
> Assignee: Esteban Gutierrez
> Attachments:
> 0001-HBASE-16774-shell-Add-coverage-to-TestShell-when-Zoo.patch,
> 0001-HBASE-16774-shell-Add-coverage-to-TestShell-when-Zoo.patch,
> HBASE-16774.001.patch
>
>
> While testing a couple of things in master I noticed that after some of the
> changes done in HBASE-16117 the hbase shell would die when there is no
> ZooKeeper server up or if we get another ZK exception. This is to add
> coverage to test the shell when ZK is not up or if we get another exception.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)