Hi,
Found one Java-level deadlock:
=
SocketListener0-7:
waiting to lock monitor 0x0845e1fc (object 0x54f95838, a
org.apache.hadoop.mapred.JobTracker),
which is held by IPC Server handler 0 on 54311
IPC Server handler 0 on 54311:
waiting to lock monitor 0x4d671064
Hello,
We have two HOD questions:
(1) For our current Torque PBS setup, the number of nodes requested by
HOD (-l nodes=X) corresponds to the number of CPUs allocated, however
these nodes can be spread across various partially or empty nodes.
Unfortunately, HOD does not appear to honour the
I've been trying to trouble shoot an OOME we've been having.
When we run the job over a dataset that about 700GB (~9000 files) or larger
we will get an OOME on the map jobs. However if we run the job over smaller
set of the data then everything works out fine. So my question is: What
changes in
Sagar Naik wrote:
Hi ,
I would like to know what happens in case of DiskFull on a datanode
Does the datanode acts as block server only ?
Yes. I think so.
Does it rejects anymore Block creation request OR Namenode does not list
it for new blocks
yes. NN will not allocate it any more
Hello all,
I'd like to take the datanode's capability to handle multiple
directories to a somewhat-extreme, and get feedback on how well this
might work.
We have a few large RAID servers (12 to 48 disks) which we'd like to
transition to Hadoop. I'd like to mount each of the disks
Very cool stuff, but I don't see a reference anywhere to the author of the
visualization, which seems like poor form for a marketing video. I apologize
if I missed a reference somewhere.
Michael Ogawa at UC Davis wrote the code to generate that visualization and
open sourced it at
Very cool stuff, but I don't see a reference anywhere to the author
of the
visualization, which seems like poor form for a marketing video. I
apologize
if I missed a reference somewhere.
Jeff, you missed it!
It is the first text screen at the end of the video.
It is actually a cool open
Owen O'Malley wrote:
It is interesting, but it would be more interesting to track the
authors of the patch rather than the committer. The two are rarely
the same.
Indeed. There was a period of over a year where I wrote hardly
anything but committed almost everything. So I am vastly
Ha, that's what I get for my short attention span. Rad stuff, sorry for
missing the recognition of the code_swarm utility.
On Wed, Dec 17, 2008 at 12:38 PM, Stefan Groschupf s...@101tec.com wrote:
Very cool stuff, but I don't see a reference anywhere to the author of the
visualization, which
Hi,
I've set up a Hadoop cluster but have a problem where multiple
datanodes and tasks stop responding. I first ran into the problem
using 0.19.0, but I also see the problem at 0.18.2. Java version is
1.6.0_11.
Looking at the logs, the first sign of trouble seems to be either
If you may have turned on ipv6 on your hadoop cluster, it may cause severe
performance hit!
When I ran the gridmix2 benchmark on a newly constructed cluster, it took
30% more time than the baseline time that was obtained on a similar cluster.
I noticed that some task processes on some machines
I see createBlockException and Abandoning block quite often
When I check the datanode, they are running. I can browse file system
from that datanode:50075
However, I also notice tht a du forked off from the DN. This 'du' run
anywhere from 6mins to 30 mins.
During this time no logs are
Hey Sagar,
If the 'du' is in the D state, then that probably means bad things
for your hardware.
I recommend looking in dmesg and /var/log/messages for anything
interesting, as well as perform a hard-drive diagnostic test (may be
as simple as a SMART tests) to see if there's an issue.
NullWritable has a get() method that returns the singleton instance of the
NullWritable.
- Aaron
On Tue, Dec 16, 2008 at 9:30 AM, David Coe david@chalklabs.net wrote:
Owen O'Malley wrote:
On Dec 16, 2008, at 9:14 AM, David Coe wrote:
Does the SequenceFileOutputFormat work with
Brian Bockelman wrote:
Hey Sagar,
If the 'du' is in the D state, then that probably means bad things
for your hardware.
I recommend looking in dmesg and /var/log/messages for anything
interesting, as well as perform a hard-drive diagnostic test (may be
as simple as a SMART tests) to see if
Hi All:
I am setting up 2 grids, each with its own HDFS. The grids are unaware of each
other but exist on the same network.
I'd like to copy data from one HDFS to the other. Is there a way to do this
simply, or do I need to cobble together scripts to copy from HDFS on one side
and pipe to a
try
hadoop distcp
more info here
http://hadoop.apache.org/core/docs/current/distcp.html
Documentation is for current release, but looking hadoop distcp should print
out help message.
Thanks,
Lohit
- Original Message
From: C G parallel...@yahoo.com
To: core-user@hadoop.apache.org
Craig,
Hello,
We have two HOD questions:
(1) For our current Torque PBS setup, the number of nodes requested by
HOD (-l nodes=X) corresponds to the number of CPUs allocated, however
these nodes can be spread across various partially or empty nodes.
Unfortunately, HOD does not appear to
On Dec 16, 2008, at 9:30 AM, David Coe wrote:
Since the SequenceFileOutputFormat doesn't like nulls, how would I use
NullWritable? Obviously output.collect(key, null) isn't working.
If I
change it to output.collect(key, new IntWritable()) I get the result I
want (plus an int that I
19 matches
Mail list logo