[
https://issues.apache.org/jira/browse/CURATOR-258?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14877339#comment-14877339
]
ASF GitHub Bot commented on CURATOR-258:
----------------------------------------
Github user Randgalt commented on a diff in the pull request:
https://github.com/apache/curator/pull/102#discussion_r39924798
--- Diff:
curator-recipes/src/main/java/org/apache/curator/framework/recipes/nodes/PersistentEphemeralNode.java
---
@@ -74,10 +74,18 @@
@Override
public void process(WatchedEvent event)
{
- if ( event.getType() == EventType.NodeDeleted)
- {
- createNode();
- }
+ if ( event.getType() == EventType.NodeDeleted)
+ {
+ createNode();
+ }
+ else if ( event.getType() == EventType.NodeDataChanged)
+ {
+ try {
+ watchNode();
+ } catch (Exception e) {
+ log.error(String.format("Unexpected error during
watching of path: %s", basePath), e);
--- End diff --
Maybe switch to a CuratorWatcher and let it handle the exception.
> PersistentEphemeralNode stops watching after first WatchedEvent
> ---------------------------------------------------------------
>
> Key: CURATOR-258
> URL: https://issues.apache.org/jira/browse/CURATOR-258
> Project: Apache Curator
> Issue Type: Bug
> Components: Recipes
> Affects Versions: 3.0.0, 2.8.0, 2.9.0, 2.9.1
> Reporter: Njal Karevoll
>
> The `Watcher` in the `PersistentEphemeralNode` recipe only reacts to the
> `NodeDeleted` event. But it's possible to receive a `NodeDataChanged` event
> as well due to
> 1. The node has been created before the recipe has been started, so the
> first event it receives is it's own update.
> 2. The node has been created by another session, and the recipe accepts
> updating nodes belonging to a different session. In this case, it also
> receives it's own update in the `Watcher`, and no longer watches the node
> when the `NodeDeleted` event comes through (after the session of the other
> client expires). This is easily reproducible by restarting a service using a
> the recipe when the session timeout is higher than the service restart time.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
