-- Forwarded message --
From: air
Date: 2011/8/11
Subject: hadoop submit job very slow
To: CDH Users
I am using CDH3 update1 hadoop 0.20.2, the cluster is composed of 9
nodes(1NN + 8 (DN and TT)) , I found that it's very slow to submit job to
the jobtracker (using hive or hado
-- Forwarded message --
From: air
Date: 2011/8/11
Subject: Re: hadoop submit job very slow
To: CDH Users
when I execute : hadoop fsck / , it shows
Total size:2246829645313 B (Total open files size: 1144403881 B)
Total dirs:22494
Total files: 47853 (Files currently
Good morning,
I would like to store some files in the distributed cache, in order to be
opened and read from the mappers.
The files are produced by an other Job and are sequence files.
I am not sure if that format is proper for the distributed cache, as the files
in distr.cache are stored and re
Hi Sofia,
I assume that output of first job is stored on HDFS. In that case I would
directly read file from Mappers without using distributed cache. If you put
file into distributed cache that would add one more copy operation into your
process.
Thanks,
dino
On Fri, Aug 12, 2011 at 9:53 AM, Sof
Thank you for the reply!
In each map(), I need to open-read-close these files (more than 2 in the
general case, and maybe up to 20 or more), in order to make some checks.
Considering the huge amount of data in the input, making all these file
operations on HDFS will kill the performance!!! So I
You can use any kind of format for files in the distributed cache, so
yes you can use sequence files. They should be faster to parse than
most text formats.
-Joey
On Fri, Aug 12, 2011 at 4:56 AM, Sofia Georgiakaki
wrote:
> Thank you for the reply!
> In each map(), I need to open-read-close these
If you are looking for performance gains, then possibly reading these files
once during the setup() call in your Mapper and storing them in some data
structure like a Map or a List will give you benefits. Having to open/close
the files during each map call will have a lot of unneeded I/O.
Yo
This whole thread doesn't make a lot of sense.
If your first m/r job creates the sequence files, which you then use as input
files to your second job, you don't need to use distributed cache since the
output of the first m/r job is going to be in HDFS.
(Dino is correct on that account.)
Sofia
Hi All,
I'm trying to decommission data node from my cluster. I put the data node in
the /usr/lib/hadoop/conf/dfs.hosts.exclude list and restarted the name nodes.
The under-replicated blocks are starting to replicate, but it's going down in a
very slow pace. For 1 TB of data it takes over 1
Sofia correct me if I am wrong, but Mike I think this thread was about using
the output of a previous job, in this case already in sequence file format, as
in memory join data for another job.
Side note: does anyone know what the rule of thumb on file size is when using
the distributed cache vs
The balancer bandwidth setting does not affect decommissioning nodes.
Decommisssioning nodes replicate as fast as the cluster is capable.
The replication pace has many variables.
Number nodes that are participating in the replication.
The amount of network bandwidth each has.
The amount of
On Fri, Aug 12, 2011 at 11:58 AM, wrote:
> Hi All,
>
> I'm trying to decommission data node from my cluster. I put the data node
> in the /usr/lib/hadoop/conf/dfs.hosts.exclude list and restarted the name
> nodes. The under-replicated blocks are starting to replicate, but it's
> going down in a
You can configure the undocumented variable dfs.max-repl-streams to
increase the number of replications a data-node is allowed to handle
at one time. The default value is 2. [1]
-Joey
[1]
https://issues.apache.org/jira/browse/HADOOP-2606?focusedCommentId=12578700&page=com.atlassian.jira.plugin.s
I did have these settings on all the hdfs-site.xml nodes:
dfs.balance.bandwidthPerSec
131072000
dfs.max-repl-streams
50
It is still taking over 1 day or longer for 1TB of under replicated blocks to
replicate.
Thanks!
Jonathan
-Original Message-
From: Joey Echeverria [mai
It could be that your process has hung cause a particular resident
block (file) requires a very large replication factor, and your
remaining # of nodes is less than that value. This is a genuine reason
for hang (but must be fixed). The process usually waits until there
are no under-replicated block
I have a large number of gzipped web server logs on NFS that I need to pull
into HDFS for analysis by MapReduce. What is the most efficient way to do
this?
It seems like what I should do is:
hadoop fs -copyFromLocal *.gz /my/HDFS/directory
A couple of questions:
1. Is this single process, o
Just a thought...
Really quick and dirty thing to do is to turn off the node.
Within 10 minutes the node looks down to the JT and NN so it gets marked as
down.
Run an fsck and it will show the files as under replicated and then will do the
replication at the faster speed to rebalance the clust
On Fri, Aug 12, 2011 at 1:29 PM, W.P. McNeill wrote:
> I have a large number of gzipped web server logs on NFS that I need to pull
> into HDFS for analysis by MapReduce. What is the most efficient way to do
> this?
>
> It seems like what I should do is:
>
> hadoop fs -copyFromLocal *.gz /my/HDFS
I am getting this error with a mostly out of the box configuration from
version 0.20.203.0
When I try to run the wordcount examples.
$ hadoop jar hadoop-examples-0.20.203.0.jar wordcount
/user/hduser/gutenberg /user/hduser/gutenberg-output6
2011-08-12 15:45:38,299 WARN org.apache.hadoop.mapred.
I'm building a Hadoop project using Maven. I want to add
Maven dependencies to my project. What do I do?
I think the answer is I add a section to my .POM
file, but I'm not sure what the contents of this section (groupId,
artifactId etc.) should be. Googling does not turn up a clear answer. Is
the
Pre-0.21 (sustaining releases, large-scale tested) hadoop:
org.apache.hadoop
hadoop-core
0.20.203.0
Pre-0.23 (small scale tested) hadoop:
org.apache.hadoop
hadoop-mapred
...
Trunk (currently targeting 0.23.0, large-scale tested) hadoop WILL be:
org.apache.hadoop
hadoop-mapre
I want the latest version of Hadoop (with the new API). I guess that's the
trunk version, but I don't see the hadoop-mapreduce artifact listed on
https://repository.apache.org/index.html#nexus-search;quick~hadoop
On Fri, Aug 12, 2011 at 2:47 PM, Luke Lu wrote:
> Pre-0.21 (sustaining releases, la
Hi all,
I was interested in learning from how Hadoop implements their sort algorithm
in the map/reduce framework. Could someone point me to the directory of the
source code that has the mapper/reducer that the Sort example uses by
default when I invoke:
$ hadoop jar hadoop-*-examples.jar sort inp
Hello Mates:
Thanks to everyone for their help so far. I have learnt a lot and have now done
single and pseudo mode. I have a hadoop cluster but I ran jps on the master
node and slave node but not all process are started
master:
22160 NameNode
22716 Jps
22458 JobTracker
slave:
32195 Jps
I als
I'm getting this "Incompatible buildVersion" error even though they look to be
the same build version (see below). The only difference is that the tasktracker
is running on Debian-6.0 and the jobtracker on Centos-5.3. Are there known
problems with this setup?
2011-08-12 23:27:10,362 WARN mapre
Sean,
The sort impl is spread out over many files. I'd start with MapTask and
ReduceTask and follow from there on.
LMK if you need more info.
thanks,
Arun
On Aug 12, 2011, at 12:48 PM, Sean Hogan wrote:
> Hi all,
>
> I was interested in learning from how Hadoop implements their sort algori
There is a reason I capitalized WILL (SHALL) :) The current trunk
mapreduce code is influx. Once mr2 (MAPREDUCE-279) is merged into
trunk (soon!). We'll be producing hadoop-mapreduce-0.23.0-SNAPSHOT,
which depends on hadoop-hdfs-0.23.0-SNAPSHOT, which depends on
hadoop-common-0.23.0-SNAPSHOT.
If
Hello:
I did more test again but now I noticed that only 3 nodes have datanodes while
the others do not. I ran the admin report tool and the result is below. Where
do i configure the capacity?
bin/hadoop dfsadmin -report
Configured Capacity: 0 (0 KB)
Present Capacity: 0 (0 KB)
DFS Remaining:
Hi All,
Recently I was working with the CompositeInputFormat for MapSide Join and then
tried MultipleInputs for ReduceSide join. The problem I found out was that both
of these methods needed you to use JobConf class (@deprecated) and the
implementation hasnt been provided in the 0.20.203 versio
Hello:
After doing some more changes, the report shows only one datanode but different
from the ones that I selected. The problem seems to be the datanode but I am
not sure why. Does hadoop need to run as root? What if a shared filesystem is
used for the nodes, how can I specify that the data
OK, that wasn't the real error, it looks like this was:
When working with cygwin. I am guessing that the task failed. Is
mapred/task runner launch a new jvm process? That seems to be failing
MapAttempt TASK_TYPE="SETUP" TASKID="task_201108130149_0001_m_02"
TASK_ATTEMPT_ID="attempt_2011081
31 matches
Mail list logo