[jira] [Commented] (CASSANDRA-13272) "nodetool bootstrap resume" does not exit
[
https://issues.apache.org/jira/browse/CASSANDRA-13272?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16087112#comment-16087112
]
Benjamin Lerer commented on CASSANDRA-13272:
Thanks for the patch. I ran CI on the 2.2 branch for extra safety and the
failing tests are unrelated.
+1
> "nodetool bootstrap resume" does not exit
> -
>
> Key: CASSANDRA-13272
> URL: https://issues.apache.org/jira/browse/CASSANDRA-13272
> Project: Cassandra
> Issue Type: Bug
> Components: Lifecycle, Streaming and Messaging
>Reporter: Tom van der Woerdt
>Assignee: Tim Lamballais
> Labels: lhf
>
> I have a script that calls "nodetool bootstrap resume" after a failed join
> (in my environment some streams sometimes fail due to mis-tuning of stream
> bandwidth settings). However, if the streams fail again, nodetool won't exit.
> Last lines before it just hangs forever :
> {noformat}
> [2017-02-26 07:02:42,287] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12670-big-Data.db
> (progress: 1112%)
> [2017-02-26 07:02:42,287] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12670-big-Data.db
> (progress: 1112%)
> [2017-02-26 07:02:59,843] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12671-big-Data.db
> (progress: 1112%)
> [2017-02-26 09:25:51,000] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:33:45,017] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:39:27,216] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:53:33,084] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:55:07,115] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 10:06:49,557] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 10:40:55,880] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 11:09:21,025] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 12:44:35,755] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 12:49:18,867] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 13:23:50,611] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 13:23:50,612] Stream failed
> {noformat}
> At that point ("Stream failed") I would expect nodetool to exit with a
> non-zero exit code. Instead, it just wants me to ^C it.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-13272) "nodetool bootstrap resume" does not exit
[
https://issues.apache.org/jira/browse/CASSANDRA-13272?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16080213#comment-16080213
]
Tim Lamballais commented on CASSANDRA-13272:
{noformat}
>From 1fd3c4eef678eda01f1f33c95760c1976675455f Mon Sep 17 00:00:00 2001
From: Timothy George Lamballais Tessensohn
Date: Mon, 10 Jul 2017 13:47:19 +0200
Subject: [PATCH] fix NPE when stream fails
https://issues.apache.org/jira/browse/CASSANDRA-13272
---
src/java/org/apache/cassandra/service/StorageService.java | 7 +--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/src/java/org/apache/cassandra/service/StorageService.java
b/src/java/org/apache/cassandra/service/StorageService.java
index 143b402d78..919df0cda3 100644
--- a/src/java/org/apache/cassandra/service/StorageService.java
+++ b/src/java/org/apache/cassandra/service/StorageService.java
@@ -1287,8 +1287,11 @@ public class StorageService extends
NotificationBroadcasterSupport implements IE
@Override
public void onFailure(Throwable e)
{
-String message = "Error during bootstrap: " +
e.getCause().getMessage();
-logger.error(message, e.getCause());
+String message = e.getMessage();
+logger.error(message, e);
+if (e instanceof ExecutionException && e.getCause() !=
null) {
+message = e.getCause().getMessage();
+}
progressSupport.progress("bootstrap", new
ProgressEvent(ProgressEventType.ERROR, 1, 1, message));
progressSupport.progress("bootstrap", new
ProgressEvent(ProgressEventType.COMPLETE, 1, 1, "Resume bootstrap complete"));
}
{noformat}
> "nodetool bootstrap resume" does not exit
> -
>
> Key: CASSANDRA-13272
> URL: https://issues.apache.org/jira/browse/CASSANDRA-13272
> Project: Cassandra
> Issue Type: Bug
> Components: Lifecycle, Streaming and Messaging
>Reporter: Tom van der Woerdt
>Assignee: Tim Lamballais
> Labels: lhf
>
> I have a script that calls "nodetool bootstrap resume" after a failed join
> (in my environment some streams sometimes fail due to mis-tuning of stream
> bandwidth settings). However, if the streams fail again, nodetool won't exit.
> Last lines before it just hangs forever :
> {noformat}
> [2017-02-26 07:02:42,287] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12670-big-Data.db
> (progress: 1112%)
> [2017-02-26 07:02:42,287] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12670-big-Data.db
> (progress: 1112%)
> [2017-02-26 07:02:59,843] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12671-big-Data.db
> (progress: 1112%)
> [2017-02-26 09:25:51,000] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:33:45,017] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:39:27,216] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:53:33,084] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:55:07,115] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 10:06:49,557] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 10:40:55,880] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 11:09:21,025] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 12:44:35,755] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 12:49:18,867] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 13:23:50,611] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 13:23:50,612] Stream failed
> {noformat}
> At that point ("Stream failed") I would expect nodetool to exit with a
> non-zero exit code. Instead, it just wants me to ^C it.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-13272) "nodetool bootstrap resume" does not exit
[
https://issues.apache.org/jira/browse/CASSANDRA-13272?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16080131#comment-16080131
]
Tim Lamballais commented on CASSANDRA-13272:
[~blerer], I'll update it to retrieve the cause in case of ExecutionException.
It'll then pass that message back to the client, but I'll use the original
throwable in the call to log.error, so we don't lose the original stacktrace in
the logs.
Cheerio,
Tim
> "nodetool bootstrap resume" does not exit
> -
>
> Key: CASSANDRA-13272
> URL: https://issues.apache.org/jira/browse/CASSANDRA-13272
> Project: Cassandra
> Issue Type: Bug
> Components: Lifecycle, Streaming and Messaging
>Reporter: Tom van der Woerdt
>Assignee: Tim Lamballais
> Labels: lhf
>
> I have a script that calls "nodetool bootstrap resume" after a failed join
> (in my environment some streams sometimes fail due to mis-tuning of stream
> bandwidth settings). However, if the streams fail again, nodetool won't exit.
> Last lines before it just hangs forever :
> {noformat}
> [2017-02-26 07:02:42,287] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12670-big-Data.db
> (progress: 1112%)
> [2017-02-26 07:02:42,287] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12670-big-Data.db
> (progress: 1112%)
> [2017-02-26 07:02:59,843] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12671-big-Data.db
> (progress: 1112%)
> [2017-02-26 09:25:51,000] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:33:45,017] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:39:27,216] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:53:33,084] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:55:07,115] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 10:06:49,557] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 10:40:55,880] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 11:09:21,025] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 12:44:35,755] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 12:49:18,867] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 13:23:50,611] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 13:23:50,612] Stream failed
> {noformat}
> At that point ("Stream failed") I would expect nodetool to exit with a
> non-zero exit code. Instead, it just wants me to ^C it.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-13272) "nodetool bootstrap resume" does not exit
[
https://issues.apache.org/jira/browse/CASSANDRA-13272?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16080063#comment-16080063
]
Benjamin Lerer commented on CASSANDRA-13272:
bq. I think at the time I thought {{Throwable e}} is usually
{{ExecutionException}} so I wanted to get what caused that execution error.
In this case, we should probably retrieve the cause only if {{e}} is an
{{ExecutionException}} and the cause is not {{null}}. It should never be the
case if {{e}} is an {{ExecutionException}} but better be safe than sorry.
[~Wimtie] could you update your patch?
> "nodetool bootstrap resume" does not exit
> -
>
> Key: CASSANDRA-13272
> URL: https://issues.apache.org/jira/browse/CASSANDRA-13272
> Project: Cassandra
> Issue Type: Bug
> Components: Lifecycle, Streaming and Messaging
>Reporter: Tom van der Woerdt
>Assignee: Tim Lamballais
> Labels: lhf
>
> I have a script that calls "nodetool bootstrap resume" after a failed join
> (in my environment some streams sometimes fail due to mis-tuning of stream
> bandwidth settings). However, if the streams fail again, nodetool won't exit.
> Last lines before it just hangs forever :
> {noformat}
> [2017-02-26 07:02:42,287] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12670-big-Data.db
> (progress: 1112%)
> [2017-02-26 07:02:42,287] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12670-big-Data.db
> (progress: 1112%)
> [2017-02-26 07:02:59,843] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12671-big-Data.db
> (progress: 1112%)
> [2017-02-26 09:25:51,000] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:33:45,017] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:39:27,216] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:53:33,084] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:55:07,115] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 10:06:49,557] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 10:40:55,880] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 11:09:21,025] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 12:44:35,755] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 12:49:18,867] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 13:23:50,611] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 13:23:50,612] Stream failed
> {noformat}
> At that point ("Stream failed") I would expect nodetool to exit with a
> non-zero exit code. Instead, it just wants me to ^C it.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-13272) "nodetool bootstrap resume" does not exit
[
https://issues.apache.org/jira/browse/CASSANDRA-13272?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16077675#comment-16077675
]
Yuki Morishita commented on CASSANDRA-13272:
I think at the time I thought {{Throwable e}} is usually {{ExecutionException}}
so I wanted to get what caused that execution error.
I'm not sure what's causing its cause to be null, but it looks it can happen.
I think we need proper null handling there.
> "nodetool bootstrap resume" does not exit
> -
>
> Key: CASSANDRA-13272
> URL: https://issues.apache.org/jira/browse/CASSANDRA-13272
> Project: Cassandra
> Issue Type: Bug
> Components: Lifecycle, Streaming and Messaging
>Reporter: Tom van der Woerdt
>Assignee: Tim Lamballais
> Labels: lhf
>
> I have a script that calls "nodetool bootstrap resume" after a failed join
> (in my environment some streams sometimes fail due to mis-tuning of stream
> bandwidth settings). However, if the streams fail again, nodetool won't exit.
> Last lines before it just hangs forever :
> {noformat}
> [2017-02-26 07:02:42,287] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12670-big-Data.db
> (progress: 1112%)
> [2017-02-26 07:02:42,287] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12670-big-Data.db
> (progress: 1112%)
> [2017-02-26 07:02:59,843] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12671-big-Data.db
> (progress: 1112%)
> [2017-02-26 09:25:51,000] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:33:45,017] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:39:27,216] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:53:33,084] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:55:07,115] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 10:06:49,557] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 10:40:55,880] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 11:09:21,025] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 12:44:35,755] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 12:49:18,867] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 13:23:50,611] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 13:23:50,612] Stream failed
> {noformat}
> At that point ("Stream failed") I would expect nodetool to exit with a
> non-zero exit code. Instead, it just wants me to ^C it.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-13272) "nodetool bootstrap resume" does not exit
[
https://issues.apache.org/jira/browse/CASSANDRA-13272?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16076144#comment-16076144
]
Benjamin Lerer commented on CASSANDRA-13272:
I had a look at the sources and I am not sure to understand the purpose of the
original code. Everywhere else in {{StorageService}} the full chain of
Exceptions is being logged.
If the wrapping Exception was created without a message it will automatically
use the message of the wrapped exception. Due to that, we should probably use:
{code}
@Override
public void onFailure(Throwable e)
{
String message = "Error during bootstrap: " +
e.getMessage();
logger.error(message, e);
progressSupport.progress("bootstrap", new
ProgressEvent(ProgressEventType.ERROR, 1, 1, message));
progressSupport.progress("bootstrap", new
ProgressEvent(ProgressEventType.COMPLETE, 1, 1, "Resume bootstrap complete"));
}
{code}
[~yukim] Do you remember why you used the exception cause instead of the
exception?
> "nodetool bootstrap resume" does not exit
> -
>
> Key: CASSANDRA-13272
> URL: https://issues.apache.org/jira/browse/CASSANDRA-13272
> Project: Cassandra
> Issue Type: Bug
> Components: Lifecycle, Streaming and Messaging
>Reporter: Tom van der Woerdt
>Assignee: Tim Lamballais
> Labels: lhf
>
> I have a script that calls "nodetool bootstrap resume" after a failed join
> (in my environment some streams sometimes fail due to mis-tuning of stream
> bandwidth settings). However, if the streams fail again, nodetool won't exit.
> Last lines before it just hangs forever :
> {noformat}
> [2017-02-26 07:02:42,287] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12670-big-Data.db
> (progress: 1112%)
> [2017-02-26 07:02:42,287] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12670-big-Data.db
> (progress: 1112%)
> [2017-02-26 07:02:59,843] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12671-big-Data.db
> (progress: 1112%)
> [2017-02-26 09:25:51,000] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:33:45,017] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:39:27,216] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:53:33,084] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:55:07,115] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 10:06:49,557] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 10:40:55,880] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 11:09:21,025] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 12:44:35,755] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 12:49:18,867] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 13:23:50,611] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 13:23:50,612] Stream failed
> {noformat}
> At that point ("Stream failed") I would expect nodetool to exit with a
> non-zero exit code. Instead, it just wants me to ^C it.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-13272) "nodetool bootstrap resume" does not exit
[
https://issues.apache.org/jira/browse/CASSANDRA-13272?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16074990#comment-16074990
]
Benjamin Lerer commented on CASSANDRA-13272:
I just realize that your patch change the current behavior. The cause should be
the cause of the Exception or the Exception itself not the root cause of the
all chain.
> "nodetool bootstrap resume" does not exit
> -
>
> Key: CASSANDRA-13272
> URL: https://issues.apache.org/jira/browse/CASSANDRA-13272
> Project: Cassandra
> Issue Type: Bug
> Components: Lifecycle, Streaming and Messaging
>Reporter: Tom van der Woerdt
>Assignee: Tim Lamballais
> Labels: lhf
>
> I have a script that calls "nodetool bootstrap resume" after a failed join
> (in my environment some streams sometimes fail due to mis-tuning of stream
> bandwidth settings). However, if the streams fail again, nodetool won't exit.
> Last lines before it just hangs forever :
> {noformat}
> [2017-02-26 07:02:42,287] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12670-big-Data.db
> (progress: 1112%)
> [2017-02-26 07:02:42,287] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12670-big-Data.db
> (progress: 1112%)
> [2017-02-26 07:02:59,843] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12671-big-Data.db
> (progress: 1112%)
> [2017-02-26 09:25:51,000] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:33:45,017] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:39:27,216] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:53:33,084] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:55:07,115] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 10:06:49,557] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 10:40:55,880] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 11:09:21,025] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 12:44:35,755] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 12:49:18,867] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 13:23:50,611] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 13:23:50,612] Stream failed
> {noformat}
> At that point ("Stream failed") I would expect nodetool to exit with a
> non-zero exit code. Instead, it just wants me to ^C it.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-13272) "nodetool bootstrap resume" does not exit
[
https://issues.apache.org/jira/browse/CASSANDRA-13272?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16074970#comment-16074970
]
Benjamin Lerer commented on CASSANDRA-13272:
Thanks for the patch.
I guess, based on the patch, that the problem came from the fact that
{{e.getCause().getMessage()}} was causing a {{NPE}}?
Regarding the patch, could you add some javadoc and a unit test for
{{Throwables::getRootCause}}.
Could you also attach the patch as a file?
> "nodetool bootstrap resume" does not exit
> -
>
> Key: CASSANDRA-13272
> URL: https://issues.apache.org/jira/browse/CASSANDRA-13272
> Project: Cassandra
> Issue Type: Bug
> Components: Lifecycle, Streaming and Messaging
>Reporter: Tom van der Woerdt
>Assignee: Tim Lamballais
> Labels: lhf
>
> I have a script that calls "nodetool bootstrap resume" after a failed join
> (in my environment some streams sometimes fail due to mis-tuning of stream
> bandwidth settings). However, if the streams fail again, nodetool won't exit.
> Last lines before it just hangs forever :
> {noformat}
> [2017-02-26 07:02:42,287] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12670-big-Data.db
> (progress: 1112%)
> [2017-02-26 07:02:42,287] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12670-big-Data.db
> (progress: 1112%)
> [2017-02-26 07:02:59,843] received file
> /var/lib/cassandra/data/keyspace/table-63d5d42009fa11e5879ebd9463bffdac/mc-12671-big-Data.db
> (progress: 1112%)
> [2017-02-26 09:25:51,000] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:33:45,017] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:39:27,216] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:53:33,084] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 09:55:07,115] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 10:06:49,557] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 10:40:55,880] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 11:09:21,025] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 12:44:35,755] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 12:49:18,867] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 13:23:50,611] session with /10.x.y.z complete (progress: 1112%)
> [2017-02-26 13:23:50,612] Stream failed
> {noformat}
> At that point ("Stream failed") I would expect nodetool to exit with a
> non-zero exit code. Instead, it just wants me to ^C it.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
