Search Discussions

83 discussions - 269 posts

  • How come running a Compute Stats on a table that is over 17TB in size with almost 2500 partitions is taking over 6 hours to run? We just upgraded to 1.2.3, and this is the first time running it. Will ...
    Benjamin KimBenjamin Kim
    Jan 12, 2014 at 3:47 am
    Feb 13, 2014 at 4:46 pm
  • Hi guys. I started to get strange error in Impala (which is installed and managed via Cloudera Manager), not even sure what might have caused it: [localhost:21000] describe services; Query: describe ...
    Dmitriy MorozovDmitriy Morozov
    Jan 2, 2014 at 7:54 pm
    Jan 3, 2014 at 5:58 pm
  • Hi Antony, You can write the query result into HDFS by issue "insert into dst_tbl <your query ". Would you mind sharing what's the end goal of your query? Do you simply need the result written to ...
    Alan ChoiAlan Choi
    Jan 20, 2014 at 5:41 pm
    Jan 24, 2014 at 2:39 pm
  • Hi, I'm using impala 1.2.3 CDH4 on CentOS 6.5 x86_64. I've built an UDF function that calculates the Base64 value of a given string: StringVal Base64Encode(FunctionContext* context, const StringVal& ...
    Sammy YuSammy Yu
    Jan 18, 2014 at 12:11 am
    Jan 29, 2014 at 3:17 am
  • My testing setup in Impala shell is 3 sessions connected to 3 different Impala nodes. I create 1 distinct table in each session. The other sessions do not see the newly created table made by other ...
    Benjamin KimBenjamin Kim
    Jan 14, 2014 at 7:15 pm
    Jan 17, 2014 at 6:38 pm
  • Hi, I'm running impala 1.2.3 on with a rcfile table with 38687 partitions that was created from hive. Afterwards, I did a refresh metadata and compared the select count(1) results and noticed that ...
    Sammy YuSammy Yu
    Jan 11, 2014 at 3:39 am
    Jan 15, 2014 at 6:15 pm
  • Hi, we are currently experimenting with partitions in Parquet tables. We have a table with column "type" and partitioned the table accoring to the distinct values of this column. Then we run some ...
    Alexander SchätzleAlexander Schätzle
    Jan 17, 2014 at 10:49 am
    Jan 17, 2014 at 6:16 pm
  • Hi all, I''m getting the following output from the profile statement for a query that joins two tables (the plan specifies the join as inner join(broadcast). The bulk of the time is "leftchildtime" ...
    Avrilia FloratouAvrilia Floratou
    Jan 27, 2014 at 9:55 pm
    Jan 29, 2014 at 9:23 pm
  • it means the catalogd can't find the com.cloudera.impala.catalog.HdfsStorageDescriptor class. you should check the classpath before you start catalogd daemon. you can print the $CLASSPATH , and check ...
    Charles DAngCharles DAng
    Jan 26, 2014 at 1:02 pm
    Jan 26, 2014 at 1:38 pm
  • Hi, this is a screenshot from the Impala Query Details page in Cloudera Manager (in German, sorry) ...
    Alexander SchätzleAlexander Schätzle
    Jan 17, 2014 at 4:22 pm
    Jan 21, 2014 at 8:50 am
  • I would like to know why querying an HBase table takes so long. If I run the same query in Hive, it takes far less time. We are trying to read 1 day's worth of event logs data. The dataset has 502M ...
    Benjamin KimBenjamin Kim
    Jan 18, 2014 at 11:27 pm
    Jan 20, 2014 at 7:55 pm
  • Performing a regular query in Impala 1.2.3-1.p0.97 select * from table where to_col like "%value%" Returns Bad status for request TFetchResultsReq(operationHandle=TOperationHandle(hasResultSet=True, ...
    Patrick o'learyPatrick o'leary
    Jan 15, 2014 at 8:28 pm
    Jan 16, 2014 at 8:18 pm
  • Is it possible to add index on Impala table stored as Parquet file? If yes - can someone put the whole example and usage? thanks, Vladimir To unsubscribe from this group and stop receiving emails ...
    Jan 9, 2014 at 1:07 am
    Jan 10, 2014 at 4:25 pm
  • I have a multi-node cloudera CDH clsuter, cloudera agent is getting timeout exception because of time difference in my cluster nodes. How to Synchronize clock time in the different nodes in cloudera ...
    Pari MarguPari Margu
    Jan 7, 2014 at 11:46 am
    Jan 7, 2014 at 5:33 pm
  • I keep getting errors with most queries on a 600M row table (about 200GB of text files). This is as part of a proof of concept using AWS and all nodes are m1.large with 7.5GB memory. I'm in the ...
    Mauricio AristizabalMauricio Aristizabal
    Jan 29, 2014 at 10:16 pm
    Mar 5, 2014 at 11:57 pm
  • DOH , figured it out Got to issue this command before I can see new tables INVALIDATE METADATA thanks sanjay To unsubscribe from this group and stop receiving emails from it, send an email to ...
    Sanjay SubramanianSanjay Subramanian
    Jan 31, 2014 at 3:22 pm
    Feb 6, 2014 at 9:40 am
  • Hi, My real-time streaming approach is the following: - have a mixed format table with an lzo compressed text journal file, the rest is in parquet - data get appended to the text file and when it ...
    György BaloghGyörgy Balogh
    Jan 23, 2014 at 11:09 am
    Jan 27, 2014 at 10:05 pm
  • Dear all, After upgrading from 1.1 to 1.2.1 (due to issue 592), though adding more than 2k partition is applicable but we notice the overall alter table responding time is getting worser. normally ...
    Jason shihJason shih
    Jan 24, 2014 at 4:41 am
    Jan 25, 2014 at 7:45 am
  • Hi Cloudera Impala Team: Now I have hadoop environment with chd5 that is installed using tar files. I prepare to setup rpm impala-1.2.0 on my RHEL-5.7,but the console prints as follows ...
    Hanks TomHanks Tom
    Jan 23, 2014 at 5:35 am
    Jan 23, 2014 at 7:18 am
  • Is it possible to have a Hive table with previous partitions be in AVRO and subsequent partitions be in Parquet going forward and still work in both Hive and Impala? If so, is it documented anywhere? ...
    Benjamin KimBenjamin Kim
    Jan 15, 2014 at 6:41 pm
    Jan 15, 2014 at 8:50 pm
  • Dear all, I tried to build Impala 1.2.3 from source on CentOS 6.3. But unlike 1.1, I got failed to build Impala. The error message is showen below ...
    Jung-Yup LeeJung-Yup Lee
    Jan 13, 2014 at 8:21 am
    Jan 15, 2014 at 9:20 am
  • I am running CDH5 beta on 3 VMs with Centos 6.5 and 8GB of RAM. I have a ~27GB set of data in CSV format in HDFS that I can't import to a parquet / snappy table without out of memory errors. I was ...
    Michael NelsonMichael Nelson
    Jan 3, 2014 at 9:36 pm
    Jan 8, 2014 at 2:04 am
  • Thanks Darren, Cloudera Manager API is something which I was looking for ! That worked! ~Ashish To unsubscribe from this group and stop receiving emails from it, send an email to ...
    Ashish AgrawalAshish Agrawal
    Jan 30, 2014 at 5:40 am
    Oct 2, 2014 at 2:56 pm
  • Hi Alan, thx for your hint. We came to the same solution but looking at the INSERT command it seems that we have to specify the PARTITION clause and the corresponding column to be the last column in ...
    Alexander SchätzleAlexander Schätzle
    Jan 14, 2014 at 10:08 am
    Sep 23, 2014 at 11:47 am
  • Hi there, I upgraded our impala cluster to the latest release 1.2.2, all the daemons started successfully. I can see the catalogs on the catalogd's debug page, but can't see that on the impalad's ...
    Zesheng WuZesheng Wu
    Jan 27, 2014 at 12:37 am
    Jan 27, 2014 at 7:02 am
  • Hi All We plan to setup a Impala cluster in AWS with 5 to 10 machines to evaluate it. It suggests the following hardware requirements in Impala installation guide: Memory - 128 GB or more ...
    Jan 26, 2014 at 7:02 am
    Jan 26, 2014 at 8:16 am
  • hi all, Do anybody meet the matter that impala happened to stop, When I queryed data by impala-shell, I found the error like that:ERROR: Couldn't open transport for clogserver-233:22000(connect() ...
    Zhong shiZhong shi
    Jan 24, 2014 at 6:11 am
    Jan 25, 2014 at 1:44 am
  • Hi all, I have some question regarding the output of the profile statement. I'm running a simple query that selects data from a table, performs an aggregation and writes the result to another table ...
    Avrilia FloratouAvrilia Floratou
    Jan 22, 2014 at 6:10 pm
    Jan 23, 2014 at 6:24 am
  • Hi Paisit, Yes, Impala support %. See this: http://www.cloudera.com/content/cloudera-content/cloudera-docs/Impala/latest/Installing-and-Using-Impala/ciiu_langref_sql.html?scroll=like_unique_1 If you ...
    Alan ChoiAlan Choi
    Jan 18, 2014 at 7:08 am
    Jan 21, 2014 at 9:30 pm
  • Hi, I'm running impala 1.2.3 with CDH4 on Amazon Web Services. I noticed that every once in a while the number of backends will drop. I took a closer look at it and it seems to be DNS related ...
    Sammy YuSammy Yu
    Jan 17, 2014 at 1:11 am
    Jan 20, 2014 at 7:10 pm
  • Hi, After I set "statestore_subscriber_timeout_seconds" to 60sec, now it's working without reconnection messages. But, there is still something wrong. I query with the following parameter, impala set ...
    Jan 18, 2014 at 2:06 pm
    Jan 19, 2014 at 3:30 am
  • When trying to drop a table, I receive the following error message: ERROR: InternalException: java.lang.RuntimeException: commitTransaction was called but openTransactionCalls = 0. This probably ...
    Jan 13, 2014 at 12:49 pm
    Jan 17, 2014 at 12:50 pm
  • Is there a setting to let Impala spill to disk when memory limits are exceeded or is this a future enhancement coming? I recently followed the guidelines stated in Setting up a Multi-tenant Cluster ...
    Benjamin KimBenjamin Kim
    Jan 13, 2014 at 5:00 pm
    Jan 15, 2014 at 8:08 pm
  • Hi all - Impala's documentation states that "For clusters running production workloads, you might load-balance between the nodes by submitting each query to a different Impala daemon in round-robin ...
    Noam CohenNoam Cohen
    Jan 15, 2014 at 5:56 pm
    Jan 15, 2014 at 7:26 pm
  • Dear guys, would there have some auto loading ability in impala? i am wondering does impala has a checking ability when new data streams flushed into hdfs (from flume), then it loads these data into ...
    Stephen AamirStephen Aamir
    Jan 10, 2014 at 2:40 am
    Jan 10, 2014 at 11:36 am
  • To Whom May Concern, As we know, to achieve the best performance of impala, we need to use parquet file and a partitioned table, supposing i have defined a partitioned table like "create table ...
    Stephen AamirStephen Aamir
    Jan 8, 2014 at 2:25 am
    Jan 9, 2014 at 2:58 am
  • Hi, after I upgraded impalad the same query is not working anymore and impalad crashes. In attached you can see the logs of the servers involved + the core dump. let me know if you need more ...
    Mario CasolaMario Casola
    Jan 8, 2014 at 5:44 am
    Jan 8, 2014 at 10:07 pm
  • Hi Alex, These frame errors will degrade your overall performance, and you should try to figure out how to fix them. I'm not sure of how these normally get fixed, but a quick google gives the ...
    Darren LoDarren Lo
    Jan 8, 2014 at 4:32 pm
    Jan 8, 2014 at 4:44 pm
  • Hi Guys, I have some issue with impala dropping databases; [dhana225:21000] show databases; Query: show databases Query finished, fetching results ... +----------------------+ ...
    Dhanasekaran AnbalaganDhanasekaran Anbalagan
    Jan 3, 2014 at 7:58 pm
    Jan 6, 2014 at 9:44 pm
  • Hi Vinay, you may be running into this issue in Impala 1.2.3: https://issues.cloudera.org/browse/IMPALA-723 Can you confirm/deny that this is the issue just to make sure you are not facing a ...
    Alex BehmAlex Behm
    Jan 2, 2014 at 6:24 pm
    Jan 3, 2014 at 7:14 am
  • Hi, According to this, http://www.cloudera.com/content/support/en/downloads/download-components/download-products/downloads-listing/connectors/cloudera-odbc-drivers.html I would like to know whether ...
    Sorawich NarkwichitSorawich Narkwichit
    Jan 24, 2014 at 1:21 pm
    Jan 31, 2014 at 1:32 am
  • Hi I was wondering is there a speculative date for the next release of impala? IMPALA-715 has basically made my system and I guess a few more a dead duck. If there isn't one imminent, is there a ...
    Patrick o'learyPatrick o'leary
    Jan 27, 2014 at 3:08 pm
    Jan 28, 2014 at 3:30 am
  • Hi Guys, I have a table with timestamp, symbol and metadata as column names. I have no control over them and we escape all over queries and DDLs with backticks but it looks like compute stats ...
    Andrew StevensonAndrew Stevenson
    Jan 27, 2014 at 10:43 pm
    Jan 27, 2014 at 11:04 pm
  • One of the features that sounds promising in Apache Shark (running over Apache Spark) is the ability to cache tables. This should potentially improve queries performance in some cases (such as ...
    Noam CohenNoam Cohen
    Jan 26, 2014 at 1:33 pm
    Jan 27, 2014 at 8:03 am
  • Hi; I've read the docs on creating UDFs. The example of loading one of the standard hive UDFs worked for me, but when I compiled my own UDF I got a ClassNotFoundException when attempting to run it ...
    Peter LancasterPeter Lancaster
    Jan 24, 2014 at 10:03 pm
    Jan 25, 2014 at 12:21 am
  • Hi Ashish, (moving to scm-user) If you're using CM, then you should restart the Impala service using the CM management interface. Here's the CM documentation regarding starting, stopping, and ...
    Matthew JacobsMatthew Jacobs
    Jan 24, 2014 at 8:19 pm
    Jan 24, 2014 at 8:51 pm
  • I have requirement where number of columns in single Impala table will be around 400 at creation and columns number expected to increase with time(may be). will there be a limitation from impala/hive ...
    iTrainer HadoopiTrainer Hadoop
    Jan 24, 2014 at 10:48 am
    Jan 24, 2014 at 8:13 pm
  • I've been using Impala 1.2.1 for a couple of weeks now and want to upgrade to 1.2.3. Doc says to do a "yum update impala-server". I get the following output from that command: Cloudera-cdh4 ...
    Jan 20, 2014 at 11:19 pm
    Jan 21, 2014 at 2:27 pm
  • Hi, I'm a bit confused about the licences. According to the many pages Impala goes under Appache Licence, but there is the Cloudera Standard Licence with many restrictions ...
    György BaloghGyörgy Balogh
    Jan 20, 2014 at 2:06 pm
    Jan 20, 2014 at 5:49 pm
  • Hi, I am still having this problem . Network is fine and Hadoop program works as usual. so, I can't do anything for this issue. Is there any way to initialize all the settings for impala ? Thanks, ...
    Jan 17, 2014 at 10:05 am
    Jan 18, 2014 at 6:25 am
Group Navigation
period‹ prev | Jan 2014 | next ›
Group Overview
groupimpala-user @

66 users for January 2014

Alex Behm: 25 posts Benjamin Kim: 22 posts Alan Choi: 16 posts Matthew Jacobs: 14 posts Alexander Schätzle: 10 posts Henry Robinson: 8 posts Sammy Yu: 8 posts Zesheng Wu: 8 posts Dmitriy Morozov: 7 posts Jim Williams: 7 posts Darren Lo: 6 posts Ishaan Joshi: 6 posts Lenni Kuff: 6 posts Nong Li: 6 posts Vladimir: 6 posts Charles Deng: 5 posts John Russell: 5 posts Noam Cohen: 5 posts Skye Wanderman-Milne: 5 posts Stephen Aamir: 4 posts
show more