Hello,

I'm setting up accumulo on a small cluster where each node has 96GB of ram
and 24 cores.  Any recommendations on what memory settings to use for the
accumulo processes, as well as what to use for the hadoop processes (e.g.
datanode, etc)?

I did a small test just to try some things standalone on a single node,
setting the accumulo processes to 2GB of ram and the HADOOP_HEAPSIZE=2000.
 While running a map reduce job with 4 workers (each allocated 1GB of RAM),
the datanode runs out of memory about 25% of the way into the job and dies.
 The job is basically building an index, iterating over data in one table
and applying mutations to another - nothing too fancy.

Since I'm dealing with a subset of data, I set the table split threshold to
128M for testing purposes, there are currently about 170 tablets so we not
dealing with a ton of data here. Might this low split threshold be a
contributing factor?

Should I increase the HADDOP_HEAPSIZE even further?  Or will that just
delay the inevitable OOM error?

The exception we are seeing is below.

ERROR org.apache.hadoop.hdfs.server.datanode.DataNode:
DatanodeRegistration(...):DataXceiveServer: Exiting due
to:java.lang.OutOfMemoryError: unable to create new native thread
        at java.lang.Thread.start0(Native Method)
        at java.lang.Thread.start(Unknown Source)
        at
org.apache.hadoop.hdfs.server.datanode.DataXceiverServer.run(DataXceiverServer.java:133)
        at java.lang.Thread.run(Unknown Source)
Thanks for your help!

Mike

2g really should be enough, it's a bit concerning. How many nodes are you
dealing with, as that could be a factor. And which version of hadoop are
you running?
On Tue, Mar 12, 2013 at 1:35 PM, Mike Hugo <[EMAIL PROTECTED]> wrote:

> Hello,
>
> I'm setting up accumulo on a small cluster where each node has 96GB of ram
> and 24 cores.  Any recommendations on what memory settings to use for the
> accumulo processes, as well as what to use for the hadoop processes (e.g.
> datanode, etc)?
>
> I did a small test just to try some things standalone on a single node,
> setting the accumulo processes to 2GB of ram and the HADOOP_HEAPSIZE=2000.
>  While running a map reduce job with 4 workers (each allocated 1GB of RAM),
> the datanode runs out of memory about 25% of the way into the job and dies.
>  The job is basically building an index, iterating over data in one table
> and applying mutations to another - nothing too fancy.
>
> Since I'm dealing with a subset of data, I set the table split threshold
> to 128M for testing purposes, there are currently about 170 tablets so we
> not dealing with a ton of data here. Might this low split threshold be a
> contributing factor?
>
> Should I increase the HADDOP_HEAPSIZE even further?  Or will that just
> delay the inevitable OOM error?
>
> The exception we are seeing is below.
>
> ERROR org.apache.hadoop.hdfs.server.datanode.DataNode:
> DatanodeRegistration(...):DataXceiveServer: Exiting due
> to:java.lang.OutOfMemoryError: unable to create new native thread
>         at java.lang.Thread.start0(Native Method)
>         at java.lang.Thread.start(Unknown Source)
>         at
> org.apache.hadoop.hdfs.server.datanode.DataXceiverServer.run(DataXceiverServer.java:133)
>         at java.lang.Thread.run(Unknown Source)
>
>
> Thanks for your help!
>
> Mike
>

Hi Mike,
  This could be related to the maximum number of processes or files allowed for your linux user. You might try bumping these values up (e.g via /etc/security/limits.conf).

-Krishmin

On Mar 12, 2013, at 1:35 PM, Mike Hugo wrote:

> Hello,
>
> I'm setting up accumulo on a small cluster where each node has 96GB of ram and 24 cores.  Any recommendations on what memory settings to use for the accumulo processes, as well as what to use for the hadoop processes (e.g. datanode, etc)?
>
> I did a small test just to try some things standalone on a single node, setting the accumulo processes to 2GB of ram and the HADOOP_HEAPSIZE=2000.  While running a map reduce job with 4 workers (each allocated 1GB of RAM), the datanode runs out of memory about 25% of the way into the job and dies.  The job is basically building an index, iterating over data in one table and applying mutations to another - nothing too fancy.
>
> Since I'm dealing with a subset of data, I set the table split threshold to 128M for testing purposes, there are currently about 170 tablets so we not dealing with a ton of data here. Might this low split threshold be a contributing factor?
>
> Should I increase the HADDOP_HEAPSIZE even further?  Or will that just delay the inevitable OOM error?
>
> The exception we are seeing is below.
>
> ERROR org.apache.hadoop.hdfs.server.datanode.DataNode: DatanodeRegistration(...):DataXceiveServer: Exiting due to:java.lang.OutOfMemoryError: unable to create new native thread
>         at java.lang.Thread.start0(Native Method)
>         at java.lang.Thread.start(Unknown Source)
>         at org.apache.hadoop.hdfs.server.datanode.DataXceiverServer.run(DataXceiverServer.java:133)
>         at java.lang.Thread.run(Unknown Source)
>
>
> Thanks for your help!
>
> Mike

Eventually it will be 4 nodes, this particular test was running on a single
node

hadoop version is 1.0.4

we already upped the limits in /etc/security/limits.conf to:

usernamehere    hard    nofile           16384

Mike
On Tue, Mar 12, 2013 at 12:49 PM, Krishmin Rai <[EMAIL PROTECTED]> wrote:

> Hi Mike,
>   This could be related to the maximum number of processes or files
> allowed for your linux user. You might try bumping these values up (e.g via
> /etc/security/limits.conf).
>
> -Krishmin
>
> On Mar 12, 2013, at 1:35 PM, Mike Hugo wrote:
>
> > Hello,
> >
> > I'm setting up accumulo on a small cluster where each node has 96GB of
> ram and 24 cores.  Any recommendations on what memory settings to use for
> the accumulo processes, as well as what to use for the hadoop processes
> (e.g. datanode, etc)?
> >
> > I did a small test just to try some things standalone on a single node,
> setting the accumulo processes to 2GB of ram and the HADOOP_HEAPSIZE=2000.
>  While running a map reduce job with 4 workers (each allocated 1GB of RAM),
> the datanode runs out of memory about 25% of the way into the job and dies.
>  The job is basically building an index, iterating over data in one table
> and applying mutations to another - nothing too fancy.
> >
> > Since I'm dealing with a subset of data, I set the table split threshold
> to 128M for testing purposes, there are currently about 170 tablets so we
> not dealing with a ton of data here. Might this low split threshold be a
> contributing factor?
> >
> > Should I increase the HADDOP_HEAPSIZE even further?  Or will that just
> delay the inevitable OOM error?
> >
> > The exception we are seeing is below.
> >
> > ERROR org.apache.hadoop.hdfs.server.datanode.DataNode:
> DatanodeRegistration(...):DataXceiveServer: Exiting due
> to:java.lang.OutOfMemoryError: unable to create new native thread
> >         at java.lang.Thread.start0(Native Method)
> >         at java.lang.Thread.start(Unknown Source)
> >         at
> org.apache.hadoop.hdfs.server.datanode.DataXceiverServer.run(DataXceiverServer.java:133)
> >         at java.lang.Thread.run(Unknown Source)
> >
> >
> > Thanks for your help!
> >
> > Mike
>
>

Have you also increased the maximum number of processes ("nproc" in the same file)? I have definitely seen this kind of error as a result of in insufficiently large process limit.

Some more details, maybe, on these pages:

http://ww2.cs.fsu.edu/~czhang/errors.html
http://incubator.apache.org/ambari/1.2.0/installing-hadoop-using-ambari/content/ambari-chap5-3-1.html

-Krishmin

On Mar 12, 2013, at 1:52 PM, Mike Hugo wrote:

> Eventually it will be 4 nodes, this particular test was running on a single node
>
> hadoop version is 1.0.4
>
> we already upped the limits in /etc/security/limits.conf to:
>
> usernamehere    hard    nofile           16384
>
> Mike
>
>
> On Tue, Mar 12, 2013 at 12:49 PM, Krishmin Rai <[EMAIL PROTECTED]> wrote:
> Hi Mike,
>   This could be related to the maximum number of processes or files allowed for your linux user. You might try bumping these values up (e.g via /etc/security/limits.conf).
>
> -Krishmin
>
> On Mar 12, 2013, at 1:35 PM, Mike Hugo wrote:
>
> > Hello,
> >
> > I'm setting up accumulo on a small cluster where each node has 96GB of ram and 24 cores.  Any recommendations on what memory settings to use for the accumulo processes, as well as what to use for the hadoop processes (e.g. datanode, etc)?
> >
> > I did a small test just to try some things standalone on a single node, setting the accumulo processes to 2GB of ram and the HADOOP_HEAPSIZE=2000.  While running a map reduce job with 4 workers (each allocated 1GB of RAM), the datanode runs out of memory about 25% of the way into the job and dies.  The job is basically building an index, iterating over data in one table and applying mutations to another - nothing too fancy.
> >
> > Since I'm dealing with a subset of data, I set the table split threshold to 128M for testing purposes, there are currently about 170 tablets so we not dealing with a ton of data here. Might this low split threshold be a contributing factor?
> >
> > Should I increase the HADDOP_HEAPSIZE even further?  Or will that just delay the inevitable OOM error?
> >
> > The exception we are seeing is below.
> >
> > ERROR org.apache.hadoop.hdfs.server.datanode.DataNode: DatanodeRegistration(...):DataXceiveServer: Exiting due to:java.lang.OutOfMemoryError: unable to create new native thread
> >         at java.lang.Thread.start0(Native Method)
> >         at java.lang.Thread.start(Unknown Source)
> >         at org.apache.hadoop.hdfs.server.datanode.DataXceiverServer.run(DataXceiverServer.java:133)
> >         at java.lang.Thread.run(Unknown Source)
> >
> >
> > Thanks for your help!
> >
> > Mike
>
>

Many thanks, Krishmin!

I had set "nofile" but not "nproc" - the references you sent were helpful.
 I had to increase the value in /etc/security/limits.d/90-nproc.conf and
now we're up and running!

Thanks again,

Mike
On Tue, Mar 12, 2013 at 1:21 PM, Krishmin Rai <[EMAIL PROTECTED]> wrote:

> Have you also increased the maximum number of processes ("nproc" in the
> same file)? I have definitely seen this kind of error as a result of in
> insufficiently large process limit.
>
> Some more details, maybe, on these pages:
>
> http://ww2.cs.fsu.edu/~czhang/errors.html
>
> http://incubator.apache.org/ambari/1.2.0/installing-hadoop-using-ambari/content/ambari-chap5-3-1.html
>
> -Krishmin
>
>
> On Mar 12, 2013, at 1:52 PM, Mike Hugo wrote:
>
> Eventually it will be 4 nodes, this particular test was running on a
> single node
>
> hadoop version is 1.0.4
>
> we already upped the limits in /etc/security/limits.conf to:
>
> usernamehere    hard    nofile           16384
>
> Mike
>
>
> On Tue, Mar 12, 2013 at 12:49 PM, Krishmin Rai <[EMAIL PROTECTED]>wrote:
>
>> Hi Mike,
>>   This could be related to the maximum number of processes or files
>> allowed for your linux user. You might try bumping these values up (e.g via
>> /etc/security/limits.conf).
>>
>> -Krishmin
>>
>> On Mar 12, 2013, at 1:35 PM, Mike Hugo wrote:
>>
>> > Hello,
>> >
>> > I'm setting up accumulo on a small cluster where each node has 96GB of
>> ram and 24 cores.  Any recommendations on what memory settings to use for
>> the accumulo processes, as well as what to use for the hadoop processes
>> (e.g. datanode, etc)?
>> >
>> > I did a small test just to try some things standalone on a single node,
>> setting the accumulo processes to 2GB of ram and the HADOOP_HEAPSIZE=2000.
>>  While running a map reduce job with 4 workers (each allocated 1GB of RAM),
>> the datanode runs out of memory about 25% of the way into the job and dies.
>>  The job is basically building an index, iterating over data in one table
>> and applying mutations to another - nothing too fancy.
>> >
>> > Since I'm dealing with a subset of data, I set the table split
>> threshold to 128M for testing purposes, there are currently about 170
>> tablets so we not dealing with a ton of data here. Might this low split
>> threshold be a contributing factor?
>> >
>> > Should I increase the HADDOP_HEAPSIZE even further?  Or will that just
>> delay the inevitable OOM error?
>> >
>> > The exception we are seeing is below.
>> >
>> > ERROR org.apache.hadoop.hdfs.server.datanode.DataNode:
>> DatanodeRegistration(...):DataXceiveServer: Exiting due
>> to:java.lang.OutOfMemoryError: unable to create new native thread
>> >         at java.lang.Thread.start0(Native Method)
>> >         at java.lang.Thread.start(Unknown Source)
>> >         at
>> org.apache.hadoop.hdfs.server.datanode.DataXceiverServer.run(DataXceiverServer.java:133)
>> >         at java.lang.Thread.run(Unknown Source)
>> >
>> >
>> > Thanks for your help!
>> >
>> > Mike
>>
>>
>
>