Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Threaded View
Accumulo >> mail # user >> Get all splits for a node [SEC=UNOFFICIAL]


Copy link to this message
-
Re: Get all splits for a node [SEC=UNOFFICIAL]
On Mon, Sep 2, 2013 at 12:47 PM, Keith Turner <[EMAIL PROTECTED]> wrote:

>
>
>
> On Sun, Sep 1, 2013 at 8:37 PM, Dickson, Matt MR <
> [EMAIL PROTECTED]> wrote:
>
>> **
>>
>> *UNOFFICIAL*
>> Is there a way to list what splits are associated with a specific node in
>> the Accumulo cluster?  I'd like to be able to see what data is going to a
>> specified IP.  Currently if we see one node with higher ingest I'm unable
>> to identify which split is causing this to then create a better split or
>> filter that data.
>>
>> I have exported the splits to a file and can see in the !METADATA table
>> there are families of loc which have value fields of IP addresses, however
>> I'm not aware of how to query the !METADATA to get all splits an IP address.
>>
>
> You could try using grep in the accumulo shell.  Something like the
> following.
>
> grep -b 0 -c loc w.x.y.z
>
> If you just want table ID 3, then you can do the following
>
> grep -b 3 -e 3< -c loc w.x.y.z
>

Realized there is a bug in the command I sent.  It could pull in table IDs
like 30, 31, etc even though you only want table ID 3.  To avoid this, use
the following command.  It adds a ';' which will prevent the inclusion of
other table ids.

grep -b 3; -e 3< -c loc w.x.y.z
>
>
>
>> Thanks in advance.
>>
>
>