Search Discussions

41 discussions - 141 posts

  • I'm looking at the feasibility of using Impala to do analytics on a couple large fact tables, but we have a star schema with slow changing dimensions, so I'm wondering how I can update those other ...
    Mauricio AristizabalMauricio Aristizabal
    Oct 4, 2013 at 7:17 am
    Oct 4, 2013 at 10:52 pm
  • Hi guys, Found a bug in inserts into parquet files. If you only insert a partial list of columns for a parquet file, the impalad daemon crashes. Here's the test case: create table foo(a string, b ...
    Oct 10, 2013 at 5:42 pm
    Oct 14, 2013 at 6:38 pm
  • We are pleased to announce the beta release of Cloudera Enterprise 5 (CDH 5 and Cloudera Manager 5). This release has both Impala and Search integrated into CDH. It also includes many new features ...
    Wendy TurnerWendy Turner
    Oct 29, 2013 at 6:26 pm
    Nov 15, 2013 at 4:33 pm
  • I have configured CDH4 security with Cloudera Manager according to this instruction ...
    C JamesC James
    Oct 11, 2013 at 3:58 pm
    Oct 15, 2013 at 1:32 pm
  • I've noticed an unexpected performance issue with impala. I expected multiple "union all" select statements to be run by child daemons in parallel, then combined. However, based on the timings, they ...
    Oct 23, 2013 at 10:57 pm
    Oct 30, 2013 at 6:12 pm
  • Thanks! It worked. impala user didnt have write permissions to a folder in HDFS. Creating the folder and assigning the ownership to impala user sort it out. Thanks, Ashish To unsubscribe from this ...
    Ashish AgrawalAshish Agrawal
    Oct 14, 2013 at 9:46 am
    Oct 16, 2013 at 9:26 am
  • Getting this error while inserting data into parquet table. Getting this error after upgrading impala to the latest version ERROR: AnalysisException: Failed to load metadata for table: default.TMP ...
    Oct 24, 2013 at 9:43 am
    Oct 29, 2013 at 8:03 pm
  • Hi, when I use Impala to query an HBase table with two WHERE conditions (connected by AND), one of the conditions seems to be ignored: select count(*) from customer_journey where customer_city is not ...
    Henrik B.Henrik B.
    Oct 17, 2013 at 1:31 pm
    Oct 24, 2013 at 10:48 pm
  • Will Impala get a BINARY data type? The use case would be to store protobufs and apply a UDF (when that feature is available) to deserialize (and query) the content. /Petter To unsubscribe from this ...
    Oct 28, 2013 at 8:36 am
    Oct 29, 2013 at 8:52 pm
  • Hi Guys, I have an Avro backed table. HIVE and the avro tools jar can read the files and IMPALA can describe the table. However selecting from the table in IMPALA causes the several deamons to crash? ...
    Andrew StevensonAndrew Stevenson
    Oct 21, 2013 at 9:28 am
    Oct 21, 2013 at 7:11 pm
  • Hi, I'm running an impala script (release 1.1.1) that creates a table (partitioned) then performs and insert overwrite. The table is successfulyl created and data insert however I get the following ...
    Andrew StevensonAndrew Stevenson
    Oct 4, 2013 at 1:19 pm
    Oct 10, 2013 at 6:40 am
  • Hi folks, I'm trying to insert data into a parquet column from an external csv table stored on hdfs. The csv files total about 8GB of memory (about the first 3rd of the wikipedia dataset from the ...
    Oct 4, 2013 at 3:17 am
    Oct 4, 2013 at 7:04 pm
  • I have an existing Impala 1.1.1 cluster (made up of, as far as I can tell, identical servers with identical environments) that works perfectly fine. However, I added 12 new nodes to the cluster ...
    Colin MarcColin Marc
    Oct 2, 2013 at 10:13 pm
    Oct 2, 2013 at 11:19 pm
  • I'd like to efficiently load data to impala located on a centralized place. Data should be stored in parquet format on the impala cluster. I would like to maximize load throughput so my approach ...
    György BaloghGyörgy Balogh
    Oct 31, 2013 at 1:49 pm
    Nov 5, 2013 at 7:15 am
  • I've been running some tests on query throughput, and the results have been different than I expected. In short, even a few concurrent queries really slows down Impala. I have a test query that takes ...
    Oct 24, 2013 at 10:30 pm
    Oct 25, 2013 at 8:23 pm
  • Hello, I have a table which contains logging data. Each log entry is associated with an user and has despite other attributes a time-stamp Thus I have n log events for each user. Now I would like to ...
    Klausen SchaefersinhoKlausen Schaefersinho
    Oct 24, 2013 at 7:21 am
    Oct 24, 2013 at 11:15 pm
  • How many nodes in your cluster and how much RAM per node? How unique is UUID (from the group by)? (on the order of how many groups do you expect from the 140M row input?) BTW, you can rewrite the ...
    Greg RahnGreg Rahn
    Oct 23, 2013 at 3:53 am
    Oct 23, 2013 at 7:22 am
  • Hi, Pls help me on how to implement CREATE TABLE AS SELECT For simple *create table t1 as select * from t2*; I can implement as Create table t1 like t2; insert into t1 as select * from t2; But how to ...
    Adline D'SilvaAdline D'Silva
    Oct 23, 2013 at 3:05 am
    Oct 23, 2013 at 3:47 am
  • I noticed that there are two ODBC drivers available (2.5.5, 2.5.0) one for Impala and one for Hive. I was under the impression that Impala uses the same protocol as Hive Server2 so what is the ...
    Surbhi chaudhrySurbhi chaudhry
    Oct 31, 2013 at 5:10 am
    Oct 31, 2013 at 10:31 pm
  • The 21050 port is the "HiveServer2" port for Impala. You must use the Hive Driver instead of the Impala Driver if you want to connect there. To unsubscribe from this group and stop receiving emails ...
    Philippe MarseillePhilippe Marseille
    Oct 29, 2013 at 8:59 pm
    Oct 30, 2013 at 4:12 am
  • We fixed a lot of issues in the 1.1.1 release but without more information, it's hard to be sure. Are you able to repro this every time? If you just run that query repeatedly, do you hit this? The ...
    Nong LiNong Li
    Oct 21, 2013 at 4:54 pm
    Oct 21, 2013 at 5:27 pm
  • Hi, Does anyone know how to configure rollover policy for Impala daemon log files being created in /var/log/impala We need to have 10 backup indexes with 10 MB file size. Thanks, Ashish To ...
    Ashish AgrawalAshish Agrawal
    Oct 14, 2013 at 9:48 am
    Oct 14, 2013 at 6:32 pm
  • Mario, Could you give us some information to help diagnose the problem? Specifically: -- Did you build Impala yourself or is this a package install ? -- The impalad/statestored logs. If you're using ...
    Ishaan JoshiIshaan Joshi
    Oct 10, 2013 at 6:38 am
    Oct 12, 2013 at 12:57 am
  • Hi, I have impala 1.1.1 installed on a 22 node cluster. I have created a table in Hive that points to a HBase table and I AM able to view the HBase table from Hive as well. But when i queried the ...
    Sachin HadoopSachin Hadoop
    Oct 9, 2013 at 3:18 pm
    Oct 10, 2013 at 7:22 am
  • You are, and Parquet 2.0 defines the layout of those index pages, and also has the option of sorted files (ie, you could really scan contiguous sections of the columns you care about, which in your ...
    Marcel KornackerMarcel Kornacker
    Oct 9, 2013 at 5:52 pm
    Oct 9, 2013 at 6:00 pm
  • Hey Rinku - I was curious if you've resolved this issue yet, if so, could you supply the fix? If not, please see Nong's advice on how to move forward with troubleshooting. Thanks -- Ricky Saltzer ...
    Ricky SaltzerRicky Saltzer
    Oct 31, 2013 at 1:40 pm
    Oct 31, 2013 at 1:40 pm
  • Hi Adline, Impala's query parser handles keywords slightly differently than Hive. Since "count" is a keyword, Impala gets confused when you use it as a column alias. To work around this, you can ...
    Lenni KuffLenni Kuff
    Oct 25, 2013 at 3:35 pm
    Oct 25, 2013 at 3:35 pm
  • Hi Manoj, Can you send me some more logging? Here's how to do it: 1. add "--vmodule=plan-fragment-executor=3" to the impalad startup parameter. If you're using CM, go to service- Impala- ...
    Alan ChoiAlan Choi
    Oct 23, 2013 at 9:36 pm
    Oct 23, 2013 at 9:36 pm
  • Can you run "describe formatted <tbl " from the hive shell? To unsubscribe from this group and stop receiving emails from it, send an email to impala-user+unsubscribe@cloudera.org.
    Nong LiNong Li
    Oct 23, 2013 at 5:51 pm
    Oct 23, 2013 at 5:51 pm
  • Hello, I have another query which stops due to memory issues: SELECT b.dif, b.counts FROM ( SELECT a.dif as dif, count(a.uuid) as counts FROM ( SELECT uuid, ceil((max(time) - min(time))/3600000) as ...
    Klausen SchaefersinhoKlausen Schaefersinho
    Oct 22, 2013 at 2:29 pm
    Oct 22, 2013 at 2:29 pm
  • Thanks, switching the union worked. Regards Andrew From: Skye Wanderman-Milne Sent: 21/10/2013 21:11 To: impala-user Subject: Re: Deamon crash with Avro backed tables. Hi Andrew, I think the problem ...
    Andrew StevensonAndrew Stevenson
    Oct 22, 2013 at 2:14 pm
    Oct 22, 2013 at 2:14 pm
  • I'm testing Impala via ODBC with PHP on a Linux system (Ubuntu 12.04 64 bit). However, I've run into an issue I can't track down. Here's the error from the PHP script. PHP Warning: odbc_connect() ...
    Oct 19, 2013 at 7:45 am
    Oct 19, 2013 at 7:45 am
  • Hello, I would like to load quite some data (30gb / 150m rows ) into impala. What is the best way to do it? I was thinking about two approaches, that would be the easiest for me: 1) Create a table ...
    Klausen SchaefersinhoKlausen Schaefersinho
    Oct 18, 2013 at 2:57 pm
    Oct 18, 2013 at 2:57 pm
  • Multiple GROUP BY clauses is not valid SQL. If you want to GROUP BY multiple columns, use the syntax: GROUP BY col1, col2, ... Thanks Lenni To unsubscribe from this group and stop receiving emails ...
    Lenni KuffLenni Kuff
    Oct 17, 2013 at 3:52 pm
    Oct 17, 2013 at 3:52 pm
  • Hi Dejan, I completely understand your dilemma. As you've correctly pointed out, you need to use the "add partition" ddl command to add a new partition into the metastore. Simply copying over files ...
    Alex BehmAlex Behm
    Oct 16, 2013 at 5:13 pm
    Oct 16, 2013 at 5:13 pm
  • *[...]So if a datanode doesn't run on a machine, is it practically useless to run an impalad daemon on it?* * * Generally, yes, as there can be no local processing done on a node that is not a ...
    Greg RahnGreg Rahn
    Oct 15, 2013 at 3:08 pm
    Oct 15, 2013 at 3:08 pm
  • I have a Hbase table with combination rowkey, such as "timestamp + user_id", each part is a 8 bytes long number. In the Impala table, I'd like to have 2 columns, the first one is key, which is the ...
    Bin YuBin Yu
    Oct 14, 2013 at 11:43 pm
    Oct 14, 2013 at 11:43 pm
  • Hi, The v1.1_branch contains the v1.1.1 source code. https://github.com/cloudera/impala/tree/v1.1_branch To unsubscribe from this group and stop receiving emails from it, send an email to ...
    Lenni KuffLenni Kuff
    Oct 12, 2013 at 5:13 am
    Oct 12, 2013 at 5:13 am
  • Hi, The Hive Metastore is the service that manages the deletion of of table data on a drop and should tell us why the data isn't being deleted. Can you send the Metastore logs? What I suspect might ...
    Lenni KuffLenni Kuff
    Oct 8, 2013 at 5:49 am
    Oct 8, 2013 at 5:49 am
  • HI, I understand how the request is served at high level. I want to understand how planner turns request into collections of plan fragments. What strategy used? -- Regards, Nishant Patel To ...
    Nishant PatelNishant Patel
    Oct 6, 2013 at 3:59 pm
    Oct 6, 2013 at 3:59 pm
  • I'm facing similiar problem. Is there any JIRA issue for this? I'm using Impala 1.1.1. Thanks in advance, Radoslaw. -- *Radosław Sypeń* Hadoop Developer/Junior Java Developer e-mail: <span ...
    Radosław SypeńRadosław Sypeń
    Oct 1, 2013 at 7:18 pm
    Oct 1, 2013 at 7:18 pm
Group Navigation
period‹ prev | Oct 2013 | next ›
Group Overview
groupimpala-user @

45 users for October 2013

Keith Simmons: 13 posts Alan Choi: 12 posts Alex Behm: 9 posts Lenni Kuff: 8 posts Greg Rahn: 7 posts Klausen Schaefersinho: 6 posts Ricky Saltzer: 6 posts Andrew Stevenson: 5 posts Marcel Kornacker: 5 posts Mauricio Aristizabal: 5 posts Ashish Agrawal: 4 posts C James: 4 posts Screenthong Lapmahapaisan: 4 posts Vikas Singh: 4 posts Adline D'Silva: 3 posts Colin Marc: 3 posts Henrik B.: 3 posts Justin Erickson: 3 posts Lhu: 3 posts Nong Li: 3 posts
show more