Hi All,
Two questions:
1) what version of hadoop does flux work with? My assumption was that it
should not matter, but I am getting the following error when I use a
hdfsbolt with flux which I assume implies some version mismatch. I don't
have this issue if build a topology straightforward that
Hi All,
I am trying to test flux module for writing template driven topologies. I
setting the topologies in localcluster mode.
1) using template simple_wordcount.yaml
2) using kafka_spout.yaml (using TestBolt)
With (1) I don't get any output form TestBolt and the topology exits.
With (2), I get
Hi All,
What would be the reason for getting this exception while running a
topology? Everything works fine for a while, and then I get this error and
the topology dies.
org.apache.zookeeper.KeeperException$ConnectionLossException:
KeeperErrorCode = ConnectionLoss for /storm/partition_6
...
...
Hi all,
Has anyone experienced a case where storm logback freezes? The topology
seems to be functioning without an issue (I can see the results, in the
destination consumers), but the storm log shows no progress. This usually
happens a couple of hours after the topology starts, and not right away.
Hi all,
I have a simple question which probably has been asked before, but I cannot
find a concrete answer. I'd appreciate your feedback.
Assume I have a spout that emit tuples which are consumed by 4 different
bolts .
Spout A --- Bolt B
--- Bolt C
--- Bolt D
Hi All,
Is there anything out there for building topologies based on templates;
that is specifying the components in a template and just have a simple
framework that builds the topology based on that template. Sorry if this
topic has been discussed before. I couldn't find anything related.
Hi All,
What could be the reasons for a topology hanging, under a somewhat heavy
load (a few hundred mb per minute)? There is no error in the logs. I am
using kafkaspout to pull data from kafka and a simple bolt to stream the
data. My spout max pending is set to 1024. My topology is running in
of the
supervisor got off.
The only way to know what is not needed is to look at zookeeper and do
what the supervisor does to determine what should be running. - Bobby
On Wednesday, March 4, 2015 11:46 AM, clay teahouse
clayteaho...@gmail.com wrote:
Is this also the case if you kill
Hi all,
How do I remove storm generated system files programmatically without
stepping over the files that I shouldn't be deleting? I need this to
cleanup the files left behind from the aborted topologies. I tried fuser
and lsof, but with no luck. For example, fuser shows stormconf.ser and
or any of
the hadoop compute nodes.
This will work for the HdfsBolt that loads default configurations from
the classpath before overriding them with any custom configurations you set
for that bolt.
- Bobby
On Thursday, February 19, 2015 10:42 AM, clay teahouse
clayteaho
Hi All,
Has anyone used HdfsBolt with hdfs in HA mode? How would you determine
which hdfs node is the active node?
thanks
Clay
On Thursday, February 19, 2015 6:47 AM, clay teahouse
clayteaho...@gmail.com wrote:
Hi All,
Has anyone used HdfsBolt with hdfs in HA mode? How would you determine
which hdfs node is the active node?
thanks
Clay
Hi All,
I have set and exported SSL_OPTS but it is not being picked up while
running storm. I need to use client certificate with the http client
connections originated from a bolt. I have set the options on the storm
command as well, but still the same issue.
thanks
Clay
Affects Versions: 0.10.0
Environment: Linux
Reporter: clay teahouse
Storm throws IndexOutOfBoundsException error once in a while, when a
bolt with shuffle grouping to is linked its predecessor bolt.especially
with a high paralleliziation hint.
This was reported
Hi All,
I emit my tuples in batches. Do I need to put the emit in a synchronized
block? The reason I am asking, I am getting the IndexOutOfBoundsException
error once in a while, especially with a high paralleliziation hint.
According to this link, it is a bug in storm, but I am using the latest
Hi,
I have a topology running in local mode. The topology uses kafkaspout and
is configured to use the external zookeeper. But when I start the topology,
I see the following:
org.apache.storm.zookeeper.ZooKeeper - Initiating client connection,
connectString=localhost:2000
I also messages like
Hi all,
In my topology, kafka spout is responsible for over 85% of the latency. I
have tried different spout max pending and played with the buffer size and
fetch size, still no luck. Any hint on how to optimize the spout? The issue
doesn't seem to be with the kafka side, as I see high
Hi,
I have a few simple questions.
1)In storm .9.x, what is the default value for the bolt num tasks?
According to the docs, the parallelism hint no longer sets the number of
tasks, but the number of executor threads.
2)What happens if the number of tasks is less than the number of threads?
Should
at 12:01 AM, clay teahouse clayteaho...@gmail.com
wrote:
Hi,
I am trying to pass some objects to the bolts through config, but I am not
having much success. These objects are hashmap and arrarylists. I am
assuming these are serializable. Any idea what could be wrong?
thanks,
Clay
Hi All,
My topology initializes a singleton containing a s set of static objects.
These objects can be complex but are static. I want all the bolts to have
access to this singleton. It seems that the bolts can access the primitives
such as strings in this singleton, but cannot access more complex
Hi,
I am trying to use storm in cluster mode. I've started nimbus and
supervisor, but when I try to run the topology, I get the error it cannot
find or load some jar in $STORM_HOME/lib. All the jars it is complaining
about do exist there. When I do storm classpath, I see the jars that it is
Hi All,
I have the following topology
spout - Bolt1 -- Bolt2
Neither bolts are async or multi-threaded. Bolt2 uses http client to make
post/put requests to a web server. Both bolts ack the tuples before exiting
the execute.
The topology runs fine for a while under a load of about 50MB/minute.
Hi All,
Why HdfsBolt doesn't retry when the hadoop node is down or not accessible
and dies brings down the topology with it too? I can catch the run time
exception and keep the topology going, but was wondering why the retry is
not built into HdfsBolt.
thank you
Clay
clay teahouse created STORM-602:
---
Summary: HdfsBolt dies when the hadoop node is not available
Key: STORM-602
URL: https://issues.apache.org/jira/browse/STORM-602
Project: Apache Storm
Issue
a thread dump of the java process when it is in hung
state. That will clearly tell you what the problem is.
For easy diagonosis set the worker to one and possibly set the number of
tasks of spout/bolt to 1.
Thanks and Regards,
Devang
On 4 Dec 2014 21:14, clay teahouse clayteaho...@gmail.com
the code in execute method of bolt B with a log
statement and check if it's still an issue.
Thanks and Regards,
Devang
On 4 Dec 2014 19:28, clay teahouse clayteaho...@gmail.com wrote:
This is a local cluster. I don't see anything interesting in the logs
that would tell me anything. I even
Hello All,
I have this configuration:
spout - Bolt A (emits tuples) - Bolt B
Bolt A emits tuples successfully but bolt B stops receiving tuples after
the first time (it never enters the execute after the first time). The
first time execution seems to be successful. Any idea what the issue could
Hello All,
I am using kafka spout that comes with storm 0.9.3 (
https://github.com/apache/storm).
I am having several different bolts consuming the same tuples from the
spout (in the same topology). These bolts process the tuples and send the
output to different destinations. I have a couple of
Hello All,
I'd appreciate your input/pointers regarding my post on join and in-blot
caching.
thanks,
Clay
On Tue, Nov 11, 2014 at 5:09 AM, clay teahouse clayteaho...@gmail.com
wrote:
Hello All,
I need to
1) Look up some values from a source and cache them in a bolt
2) Filter a second stream
Hello All,
Is it possible emit batches of tuples, as opposed to one tuple at a time?
In other word, is it possible to batch the tuples before emitting them? An
application for batching the tuples is for example for writing the tuples
to a tcp socket but not wanting to do a flush after each tuple
Hello All,
Can someone share an example of a bolt with multi stream output, with each
particular output stream going to a particular bolt?
Bolt A =stream 1 = Bolt B
Bolt A = stream 2 = Bolt C
Bolt A = stream 3 = Bolt D
thanks,
Clay
acking enabled, as its used to
coordinate batches.
On Thu, Oct 9, 2014 at 4:09 AM, clay teahouse clayteaho...@gmail.com
wrote:
Hello,
I am trying to turn off acking by settingTOPOLOGY_ACKER_EXECUTORS to 0.
But
when I do that my trident topology fails with the following error
Hello,
I am trying to turn off acking by settingTOPOLOGY_ACKER_EXECUTORS to 0. But
when I do that my trident topology fails with the following error and
subsequently the worker dies.
java.lang.RuntimeException: backtype.storm.topology.FailedException:
Received commit for different transaction
Hi,
Is there a tcp socket spout out there?
thanks,
Clay
34 matches
Mail list logo