drill-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Khurram Faraaz <kfar...@mapr.com>
Subject Re: Issue in executing query on Drill Cluster
Date Wed, 18 Oct 2017 07:00:36 GMT
Can you please see if you can access that file as the default user in both cases, because you
mention that the default user is different in both cases.

Try to do a hadoop fs -ls on that file, from both the different users and verify.


Also, can you please share the JDBC connection string that you use to connect to Drill (embedded
+ cluster), and in both cases do you connect from sqlline from the command prompt or do you
connect from Drill's web UI ?


Thanks,

Khurram


________________________________
From: Chetan Kothari <chetan.kothari@oracle.com>
Sent: Tuesday, October 17, 2017 11:53:40 PM
To: user@drill.apache.org
Subject: RE: Issue in executing query on Drill Cluster

HI Khurram



I have given permission 777 to file /datalake/replicator/testdemo2.



I am connecting to Drill with default user.

But default user is different in both cases.



Is it creating issue?

How do we fix it?



Regards

Chetan





-----Original Message-----
From: Khurram Faraaz [mailto:kfaraaz@mapr.com]
Sent: Wednesday, October 18, 2017 12:18 PM
To: user@drill.apache.org
Subject: Re: Issue in executing query on Drill Cluster



Hi Chetan,





  1.  What are the permissions to the file /datalake/replicator/testdemo2 ?

  2.  Are you connecting as the same user to Drill in both cases, (i) in embedded mode, and
(ii) in the 4 node Drillbit cluster ?





Thanks,



Khurram



________________________________

From: Chetan Kothari <HYPERLINK "mailto:chetan.kothari@oracle.com"chetan.kothari@oracle.com>

Sent: Tuesday, October 17, 2017 9:43:42 PM

To: HYPERLINK "mailto:user@drill.apache.org"user@drill.apache.org

Subject: Issue in executing query on Drill Cluster



HI All







When I am firing following query against dataset present in alluxio from embedded-drill, it
is working normally.







select REQ_HIST_REQUISITION_NUMBER, CONVERT_FROM(REQ_HIST_CREATION_DATE,'UTF8') AS REQ_HIST_CREATION_DATE
, CONVERT_FROM(REQ_HIST_END_DATE,'UTF8') AS REQ_HIST_END_DATE, CONVERT_FROM(DEPT_NAME,'UTF8')
AS DEPT_NAME ,  CONVERT_FROM(MANAGER_FIRST_NAME,'UTF8') AS MANAGER_FIRST_NAME, CONVERT_FROM(MANAGER_LAST_NAME,'UTF8')
AS MANAGER_LAST_NAME,CONVERT_FROM(DEPT_LOCATION_NUMBER,'UTF8') AS DEPT_LOCATION_NUMBER, REQUISITION_APPROVED,
REQUISITION_FILLED,REQUISITION_CANCELLED,REQUISITION_ON_HOLD, REQUISITION_SOURCING,REQUISITION_PENDING,REQUISITION_DRAFTED,REQUISITION_DELETED,REQUISITION_REJECTED
 from chetanalluxio.`./datalake/replicator/testdemo2`







But same query when I execute from 4 Node Drill Cluster, it fails with following error.



Please note that I have not enabled parquet meta-data caching feature.







org.apache.drill.common.exceptions.UserRemoteException: VALIDATION ERROR: From line 1, column
599 to line 1, column 611: Table 'chetanalluxio../datalake/replicator/testdemo2' not found
SQL Query null [Error Id: 60f552bf-3cc1-4939-807e-827dffcab11c on bdcs4node-bdcsce-1.bddcs.oracleinternalden14.nimbula.oraclecorp.com:31010].







Also I am able to browse above file present in alluxio from Drill using show files in chetanalluxio.







I am not sure why this issue is coming, given that it is working normally if execute query
using embedded-drill.



Any inputs on this will be helpful.











Regards



Chetan



Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message