[
https://issues.apache.org/jira/browse/PHOENIX-2437?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15014513#comment-15014513
]
James Taylor commented on PHOENIX-2437:
---------------------------------------
Another possible solution might be to add back the shutdown of the
ExecutorService to PhoenixDriver.close(), so it'd look like this:
{code}
@Override
public synchronized void close() throws SQLException {
try {
closeLock.writeLock().lock();
if (closed) {
return;
}
closed = true;
} finally {
closeLock.writeLock().unlock();
}
if (services != null) {
try {
services.close();
} finally {
ExecutorService executor = services.getExecutor();
// Even if something wrong happened while closing services
above, we still
// want to set it to null. Otherwise, we will end up having a
possibly non-working
// services instance.
services = null;
executor.shutdown();
}
}
}
{code}
> Build times out intermittently
> ------------------------------
>
> Key: PHOENIX-2437
> URL: https://issues.apache.org/jira/browse/PHOENIX-2437
> Project: Phoenix
> Issue Type: Bug
> Reporter: James Taylor
>
> Looks like our build is timing out still, even with my change for
> PHOENIX-2427. [~mujtabachohan] - any insight into the new failures?
> [~samarthjain] - I'm thinking we should remove the shutdown hook altogether
> and live with exceptions that can occur when a server is going down while
> connection are still being attempted to be gotten.
> WDYT?
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
