Search Discussions
-
Hi again! I am moving forward with my *Percentile* implementation. I am first trying to get the "naïve" approach working (store all the values, sort, get the percentile from the sorted array), and ...
Julien L.
Jul 15, 2014 at 10:08 am
Jul 18, 2014 at 4:13 pm -
Hi All, I'm getting the below error while trying to query a table in default database. This table was create in Impala ( create table like hive table stored as parequet table). W0630 10:58:40.682155 ...
Impala explorer
Jul 1, 2014 at 6:39 am
Jul 3, 2014 at 1:13 am -
Hi there! I am trying to define a Percentile UDA. I followed the only example I found online, of a UDA taking multiple arguments: string concatenation in the CDH5 documentation ...
Julien L.
Jul 14, 2014 at 11:23 am
Jul 30, 2014 at 1:59 pm -
Hi Matt, No I used my JAVA code using JDBC to run the SQLs in a for loop. I checked the status using the http://<impalad IP :25000/queries page . However as I mentioned before, all of them are ...
Jim Green
Jul 9, 2014 at 9:49 pm
Jul 10, 2014 at 9:44 pm -
Hi, We have been using impyla and noticed that its performance is slower than impala-shell -B -q by a factor of 50. E.g. on a SELECT statement containing 100k rows, it takes 50 seconds with impyla ...
Matt Hollingsworth
Jul 22, 2014 at 5:47 pm
Sep 5, 2014 at 6:38 pm -
The following query has been providing very inconsistent results depending on what the query optimizer is trying to do. When I first ran it one 1.4.0 I got 300k when I expected about 2m. When I ...
Sean O'Brien
Jul 18, 2014 at 11:37 pm
Jul 30, 2014 at 8:25 pm -
I am using mysql for store metadata of HIVE. I am able to query my data just fine using hive. When I type command "show tables" in impala shell, I can see the tables. But as soon as I query (e.g ...
Gaurav Kumar
Jul 30, 2014 at 9:12 am
Jul 30, 2014 at 9:48 am -
Experts, Not able to drop an impala table (parquet format with Snappy compression) after running - COMPUTE STATS on the table. Impala Version: 1.3.1-cdh5 Using Mysql as metastore. Any workarounds on ...
Venkat Ankam
Jul 7, 2014 at 4:44 pm
Jul 9, 2014 at 3:49 pm -
I'm using CDH 5.0.2, which includes Impala 1.3.1 and HBase 0.96.1.1. I created an external table named "impala_AA" in Hive shell and mapped it to a HBase table named "AA". When AA had 10,000 rows ...
Tao Xiao
Jul 6, 2014 at 12:55 am
Aug 27, 2014 at 1:10 am -
Hi I met a case in our impala production use. We have a huge parquet partition table. Any query on this table consumes a lot of time on the state "CREATED". I attached the table stats and two query ...
邱睿
Jul 30, 2014 at 6:54 am
Aug 2, 2014 at 1:06 pm -
Experts, I am always getting this error. ERRORS: Backend 3:Unknown disk id. This will negatively affect performance. Check your hdfs settings to enable block location metadata. I checked the below ...
Venkat Ankam
Jul 29, 2014 at 5:57 pm
Jul 31, 2014 at 8:16 pm -
This question is mainly for Cloudera's product team and other users who are migrating from traditional DWs. I had hoped the second Kimball Cloudera webinar would clear up how to handle Slow Changing ...
Mauricio Aristizabal
Jul 22, 2014 at 7:01 pm
Jul 28, 2014 at 6:24 pm -
Hi guys, Generally, the performance of Impala has been extremely impressive. However, as we've added more data, there have been two areas of pain I wanted to point out: 1) Slow DDL commands adding ...
Keith Simmons
Jul 23, 2014 at 7:08 pm
Jul 23, 2014 at 9:01 pm -
Hi all, We run into the issue last night. Every time executing the query, one of impalad daemons always crashes. When run it in impala-shell, the response is below. Error communicating with impalad ...
Yukinori SUDA
Jul 4, 2014 at 7:58 am
Jul 11, 2014 at 5:24 am -
Hello, I just updated cm from 4.6.2 to 4.8.3 and impala from 1.1.1 to 1.2.2. But I can not start the impala service as Service did not start successfully; not all of the required roles started: The ...
Mx wang
Jul 29, 2014 at 8:39 pm
Jul 31, 2014 at 5:45 pm -
Hi All, I have another fun potential bug. In the following query, the result column metric_1 gets -nan for some rows. Digging into the source tables there's no obvious reason for the NAN's. What's ...
Sean O'Brien
Jul 30, 2014 at 11:14 pm
Jul 31, 2014 at 12:19 am -
Hi Team, Is there a way we can export query results from impala-shell into csv and download. Please advise. Thanks PP Raja To unsubscribe from this group and stop receiving emails from it, send an ...
PP Raja
Jul 26, 2014 at 11:23 pm
Jul 28, 2014 at 7:21 pm -
Hi, I just upgraded from impala 1.2.3 to impala 1.3.1 using rpm install (yum). However, the catalog daemon is failing to start with the follwoing error: E0711 12:46:13.735695 19903 frontend.cc:100] ...
Sukhendu chakraborty
Jul 11, 2014 at 8:04 pm
Jul 18, 2014 at 10:44 pm -
Hi, I'm a Cloudera newbie so please bear with me :) We have installed Cloudera Manager etc. on a single node (VM), the details are as follows : VM : lsb_release -a LSB Version ...
Ankita singla
Jul 8, 2014 at 9:29 am
Jul 18, 2014 at 12:45 am -
Guys: I've almost got my Impala authentication woes solved, but have one more issue that's troubling me. I configure impala (1.2.4) with the command-line switches: --ldap_auth ...
Phillip Rhodes
Jul 11, 2014 at 9:28 pm
Jul 14, 2014 at 6:06 pm -
Philip, A few questions to better understand the problem: - What do your impalad principal names look like? - Are you able to successfully execute hdfs commands in your kerberized cluster (or, for ...
Ishaan Joshi
Jul 7, 2014 at 8:05 pm
Jul 8, 2014 at 6:27 pm -
hi guys, I tried to insert data into a parquet table: create table parquet_table LIKE some_table STORED AS PARQUET insert overwrite parquet_table partition(year) select * from some_table then command ...
Kang Chen
Jul 2, 2014 at 2:59 pm
Jul 7, 2014 at 3:31 pm -
Hi, I am new to Impala. I have huge amount of avro files on HDFS I am trying to use these files in Impala. What i did so far ; create table avro1(id int, time_position timestamp) STORED AS AVRO ...
Georgi Ivanov
Jul 30, 2014 at 3:35 pm
Jul 31, 2014 at 8:45 am -
Hello, Having a fresh installation of Cloudera CDH 5.0.2 with impala every query I run executes forever. I suppose it is a configuration problem, but since I didn't do any changes in the default ...
Alex Troussov
Jul 8, 2014 at 12:23 pm
Jul 23, 2014 at 3:42 pm -
Hi, I'm facing a problem: when I launch a COMPUTE STATS on a table through JDBC, and right after getting the compute stats with SHOW COLUMN STATS on the same table again through JDBC, the COMPUTE ...
Erwann Cloarec
Jul 17, 2014 at 2:09 pm
Jul 22, 2014 at 4:57 pm -
Hi Riccardo, This occurs when too many queries are running at the same time, resulting in queries being queued by Impala's admission controller [1]. If the queries are waiting in the queue for too ...
Matthew Jacobs
Jul 21, 2014 at 9:44 pm
Jul 22, 2014 at 4:43 pm -
The impala documentation notes that impala prefers to read files of 1gb in size and with a 1gb hdfs block size. Is the 1gb file size supposed to be 1gb of uncompressed data or 1gb of compressed data? ...
Slava Markeyev
Jul 2, 2014 at 6:27 pm
Jul 22, 2014 at 3:54 pm -
Hi Julien, That is not currently possible in the impala-shell, but seems like a useful enhancement. Can you file a JIRA to track this? Thanks, Lenni To unsubscribe from this group and stop receiving ...
Lenni Kuff
Jul 18, 2014 at 4:03 pm
Jul 18, 2014 at 4:34 pm -
Hi, I have CDH 5.0.2 installed recently. Under the configuration page, I can see that dfs.client.read.shortcircuit is enabled. But when I check the various hdfs-site.xml, it is False: # grep ...
Li Ming Tsai
Jul 17, 2014 at 12:58 am
Jul 17, 2014 at 5:50 pm -
Hi, I am using Cloudera4.5 (3 nodes) cluster for Hadoop and Impala1.1 for execution of queries. My namenode is having 40GB RAMand each datanodes is having 16GB RAM. *Size of my tables is as follows ...
Ankita singla
Jul 9, 2014 at 9:43 am
Jul 15, 2014 at 12:34 am -
Hi Matt, Thanks for replying. This is my test lab so I am trying to test the behavior of admin control. I am using CM5 to manage this CDH5 cluster, so I did not find " -rm_default_memory" in ...
Jim Green
Jul 9, 2014 at 7:08 pm
Jul 9, 2014 at 9:33 pm -
Hi, We are using impala 1.2.3 and have setup the startup option to prevent users from specifying limit clauses in order by: -default_query_options default_order_by_limit=10000000 However, when we use ...
Sukhendu chakraborty
Jul 2, 2014 at 12:34 am
Jul 3, 2014 at 6:56 pm -
Hi Nico - This does indeed look like a bug; the question is how it's getting triggered and if we can fashion a workaround. Are you able to share your full Llama and Impala logs? Feel free to mail ...
Henry Robinson
Jul 29, 2014 at 11:27 pm
Jul 30, 2014 at 10:05 am -
Hello, How can I set query time out in impala v1.1.1. or CM4.7. The document of impala v1.1.1 is not clear on this problem. Will it work that just put *--idle_query_timeout=60* in CM (Services Impala ...
Mx wang
Jul 28, 2014 at 4:52 pm
Jul 28, 2014 at 6:28 pm -
Hi, impala team. I'd like to report a bug. If a IllegalStateException is occured in the policyReader_ thread while reading a sentry .ini file, this thread will simply die, also stopping the scheduled ...
Jung-Yup Lee
Jul 25, 2014 at 2:29 am
Jul 25, 2014 at 5:45 am -
Hi Dom, It looks like you are hitting an Impala bug. I have filed the following JIRA to track this issue: https://issues.cloudera.org/browse/IMPALA-1095 Thanks, Lenni To unsubscribe from this group ...
Lenni Kuff
Jul 17, 2014 at 8:57 pm
Jul 17, 2014 at 10:18 pm -
Hi All, A simple drop table seems to take a pretty long time... what can I look at to better debug a performance issue? Thanks! -Sean Query Timeline: 51s379ms - Start execution: 6.145ms (6.145ms) - ...
Sean O'Brien
Jul 15, 2014 at 8:41 pm
Jul 15, 2014 at 10:22 pm -
I am looking for documentation on what is Impala's valid date ranges for a Timestamp. We have received the Application Exception : Year is out of valid range: 1400..10000 error on data that is set to ...
Jjaco4
Jul 14, 2014 at 5:30 pm
Jul 14, 2014 at 9:58 pm -
Hi， I run impala on CDH5.0.1 with namenode HA enabled (all deployed by CM). when run the statement 'create table target_table as select * from source_table', there are errors: ERROR ...
Kang Chen
Jul 12, 2014 at 4:55 am
Jul 14, 2014 at 9:27 am -
I'm trying to load data into a table with file format parquet Camillo 2014-07-03 20:15 GMT+02:00 Romain Rigaux <<span class="m_body_email_addr" title="b176ddff4d99a8970e6ea701e1126007" ...
Siaco
Jul 4, 2014 at 8:28 am
Jul 9, 2014 at 8:35 pm -
Hi - You're missing a couple of closing parentheses. The first is after ' CAST(PCM_TEST_CMP.PCM_SPECIFIC_USE_DETAIL.DRIVER_VALUE as double)' to close the 'SUM(', and the second is after ...
Henry Robinson
Jul 7, 2014 at 9:43 pm
Jul 7, 2014 at 10:02 pm -
I have a cloudera 5.5 (3 nodes)cluster. I am using Impala for my Queries. Previously the RAM on 1 machine was 40GB(master node) and other two were 2 GB(datanodes) each. It was taking long to execute ...
Ankita singla
Jul 7, 2014 at 8:05 am
Jul 7, 2014 at 6:19 pm -
Hi I have investigated Impala for some time and I still fail to understand how does Impala handle index base file such as MapFile or SortedKeyValueFile (by Avro). From first reading its look like ...
Amir Skovronik
Jul 2, 2014 at 1:35 pm
Jul 5, 2014 at 8:20 pm -
Hello all, I was just reading the source code for libwebhdfs, specifically (/hadoop-2.4.1-src/hadoop-hdfs-project/hadoop-hdfs/src/contrib/libwebhdfs/src/hdfs_web.c) and the current implementation of ...
Bruno Alvisio
Jul 30, 2014 at 3:17 pm
Jul 30, 2014 at 3:17 pm -
Hi impala noob here. I have read that hive launches mapred jobs to answer SQL queries. Does Impala do the same ? What happens if the question being asked by the SQL cannot fit into the mapred ...
ItWasntMe
Jul 30, 2014 at 1:47 pm
Jul 30, 2014 at 1:47 pm -
Actually we are working on using Impala,Tableau and Active directory with SSL encryption. I was actually trying to collate the information how actually to implement this but I really could not get ...
Sreebalineni
Jul 29, 2014 at 6:54 pm
Jul 29, 2014 at 6:54 pm -
Hi, I recently upgraded to CDH 5.1.0. Since then I am not able to see more than 20 executing queries on the queries tab of Impala service. I followed ...
Gururaj forever
Jul 25, 2014 at 4:36 am
Jul 25, 2014 at 4:36 am -
0
string columns in parquet file saved as BINARY but without UTF8 annotation, is that bug or just so?
Hi, i just tried this in newest CDH release 5.1.0 impala still save strings without UTF8 annotation, it makes trouble if we try to use other tools to access those parquet files, such as spark 1.1.0 ...Teng Qiu
Jul 22, 2014 at 3:51 pm
Jul 22, 2014 at 3:51 pm -
Great, any reason why Hive 0.13 was not included ? Regards Il giorno giovedì 17 luglio 2014 23:26:16 UTC+2, Wendy Turner ha scritto: To unsubscribe from this group and stop receiving emails from it, ...
Krusty
Jul 18, 2014 at 9:51 am
Jul 18, 2014 at 9:51 am -
I am running a complex query from Qlikview and it fails with message - ODBC read failed. Same query runs in couple of minutes from Impala-shell. Less complex Impala queries are running well from ...
Venkat Ankam
Jul 11, 2014 at 6:18 pm
Jul 11, 2014 at 6:18 pm
Group Overview
| group | impala-user |
| categories | hadoop |
| discussions | 57 |
| posts | 197 |
| users | 59 |
| website | cloudera.com |
| irc | #hadoop |
59 users for July 2014
Archives
- November 2014 (94)
- October 2014 (212)
- September 2014 (162)
- August 2014 (218)
- July 2014 (200)
- June 2014 (172)
- May 2014 (149)
- April 2014 (150)
- March 2014 (197)
- February 2014 (180)
- January 2014 (269)
- December 2013 (207)
- November 2013 (174)
- October 2013 (141)
- September 2013 (145)
- August 2013 (149)
- July 2013 (288)
- June 2013 (299)
- May 2013 (317)
- April 2013 (220)
- March 2013 (276)
- February 2013 (147)
- January 2013 (129)
- December 2012 (109)