Hi Haitao,
To help isolate, what happens if you run a different job? Also, if you
view the namenode webui or the specific datanode webui having the issue,
are there any indicators of it being down?
Regards,
Robert
On Tue, Dec 4, 2012 at 12:49 AM, panfei wrote:
> I noticed that you are using jd
I noticed that you are using jdk 1.7 , personally I prefer 1.6.x ;
if your firewall is OK, you can check you RPC service to see if it is also
OK; and test it by telnet 10.130.110.80 50020;
I suggested hive because HQL(SQL-like) is familiar to most people, and the
learning curve is smooth;
2012/1
The firewall is OK.
Well, personally I prefer Pig. And it's a big project, switching pig to hive is
not an easy way.
thanks.
Haitao Yao
yao.e...@gmail.com
weibo: @haitao_yao
Skype: haitao.yao.final
On 2012-12-4, at 下午3:14, panfei wrote:
> check your firewall settings plz. and why not use h
check your firewall settings plz. and why not use hive to do work ?
2012/12/4 Haitao Yao
> hi, all
> I's using Hadoop 1.2.0 , java version "1.7.0_05"
> When running my pig script , the worker always report this error, and the
> MR jobs run very slow.
> Increase the dfs.socket.timeout value do
hi, all
I's using Hadoop 1.2.0 , java version "1.7.0_05"
When running my pig script , the worker always report this error, and
the MR jobs run very slow.
Increase the dfs.socket.timeout value does not work. the network is ok,
telnet to 50020 port is always ok.
he
