[
https://issues.apache.org/jira/browse/HBASE-13004?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14314274#comment-14314274
]
Ted Yu commented on HBASE-13004:
--------------------------------
lgtm
minor comments:
{code}
109 * @return the boolean result of the evaluation.
110 */
111 String explainFailure() throws E;
{code}
Please update @return description.
{code}
2928 List<Pair<HRegionInfo, ServerName>> metaLocations =
2929 MetaTableAccessor.getTableRegionsAndLocations(connection,
tableName);
2930 if (getHBaseCluster().getMaster().isAlive()) {
{code}
metaLocations can be assigned inside the if block where it is used.
> Make possible to explain why HTU.waitFor fails
> ----------------------------------------------
>
> Key: HBASE-13004
> URL: https://issues.apache.org/jira/browse/HBASE-13004
> Project: HBase
> Issue Type: Bug
> Components: test
> Affects Versions: 2.0.0
> Reporter: Andrey Stepachev
> Assignee: Andrey Stepachev
> Priority: Minor
> Attachments: HBASE-13004.patch
>
>
> HBaseTestingUtility has good methods for waiting conditions, but with one
> drawback, when waitfor fails, there is no way to explain what happened
> (fail() called from internals and that is not very easy to intercept and
> explain state of failure).
> That would be great to have enhanced Predicate which can explain state after
> it failed to wait for.
> That can looks like
> {code}
> HUT.waitFor(10000, new ExplainingPredicate() {
> boolean evaluate() {
> ... some evaluation code
> }
> String explain() {
> ... generate brief explanation of what to be expected, but not happened
> }
> }
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
