I am using Standalone HBase version 0.98.8 and Phoenix 4.2.2.
There are 20 static columns in that table. Rowkey is comprised of three columns.
Each row can have 0 to 50 dynamic columns. There is only a single column family.

In the select query, I am using all the row key columns in the where clause and 
in the same order how they are defined in table.
There are around 5,00,000 records.

The query time of searching an exact row with all static+dynamic columns is 
around 3 sec as shown in SQLLine.
When I select a single column, the time drops to as less as 300ms.

It would have been great if I could share the exact schema and query. But I am 
really in hurry for a weekend to mask the schema/query before sending.
I believe this scenario must have been caught by others also. Pls let me know.



Regards
Sumanta 

-----James Taylor <[email protected]> wrote: -----
To: user <[email protected]>
From: James Taylor <[email protected]>
Date: 07/30/2015 03:13AM
Subject: Re: Poor SELECT performance in wide table

Hi Sumanta,
Would it be possible to get more detail? Phoenix/HBase version, schema, 
queries? Where are you finding the bottlenecks to be?
Thanks,
James



On Tue, Jul 28, 2015 at 10:46 PM, Sumanta Gh <[email protected]> wrote:
Hi All,
Have anyone noticed that with the increasing number of columns (static/dynamic) 
the performance of SELECT query linearly degrades even if I mention all row key 
columns in the query.

Any comments on this is appreciated.

Thanks and regards,
Sumanta 
=====-----=====-----=====
Notice: The information contained in this e-mail
message and/or attachments to it may contain 
confidential or privileged information. If you are 
not the intended recipient, any dissemination, use, 
review, distribution, printing or copying of the 
information contained in this e-mail message 
and/or attachments to it are strictly prohibited. If 
you have received this communication in error, 
please notify us by reply e-mail or telephone and 
immediately and permanently delete the message 
and any attachments. Thank you


Reply via email to