FAQ

Search Discussions

39 discussions - 145 posts

  • Hi - I have a problem that might be easy to fix... I setup an external table under my home directory in hdfs. I don't want anyone except myself and members in my group to view this table. So I put ...
    Bps cdhBps cdh
    Sep 25, 2013 at 5:09 pm
    Nov 18, 2013 at 2:09 pm
  • Hi guys, I am installing Impala (1.1.1) in *pseudo distributed mode* on ubuntu 12.04 LTS. CDH is 4.3.1 (latest) After all the installation, when i write impala-shell command it shows *Error ...
    Karam anejaKaram aneja
    Sep 5, 2013 at 3:19 pm
    Sep 16, 2013 at 2:17 pm
  • Hi everyone, I plan to build an Impala proof-of concept. I want to benchmak Impala vs Hive queries. I'm working on a star model, where the main table is about 3 billion rows and 100Go. There is no ...
    X6i4uybzX6i4uybz
    Sep 30, 2013 at 7:24 am
    Oct 5, 2013 at 2:14 am
  • Dear all, I encountered an unknown error 255 while inserting about 1 GB size of 1 year data into table partitioned by date, if works fine with 1 year data, but when I increased to 10 year(3650 days) ...
    Eric HuangEric Huang
    Sep 6, 2013 at 11:52 am
    Sep 12, 2013 at 10:06 pm
  • Hi, I've been thinking about joining against a table with ip ranges for geoip lookups e.g. create table geoip ( ip_start_int bigint, ip_end_int bigint, country string ); create table foo ( ip_int ...
    Gerrard McnultyGerrard Mcnulty
    Sep 30, 2013 at 2:16 pm
    Feb 20, 2014 at 12:05 am
  • Can you provide more detail about your install? Cloudera packages, parcels or did you build your own? Did you use Cloudera Manager to do the install? Does a basic "select * from hbase_table limit 1" ...
    Greg RahnGreg Rahn
    Sep 30, 2013 at 5:39 pm
    Oct 1, 2013 at 4:28 pm
  • Hi All, I tried searching the impala Jira / users group to see if this is a widely known issue but haven't had much luck. I'm noticing that when part of my join's constraint is a partition key and I ...
    Sean O'BrienSean O'Brien
    Sep 11, 2013 at 6:11 pm
    Oct 2, 2013 at 3:50 pm
  • Found below logs from node which failed. Tuple(id=1 size=8 slots=[Slot(id=0 type=BIGINT col=-1 offset=0 null=(offset=0 mask=0))]) I0910 02:25:15.116484 5514 mem-limit.h:86] Query: 0:0Exceeded limit ...
    Nishant PatelNishant Patel
    Sep 10, 2013 at 7:28 am
    Sep 16, 2013 at 6:15 pm
  • impala-user will probably offer better advise, since this doesn't seem CM-related. Moving to <span class="m_body_email_addr" title="4f11bca44eddf56fc4f71409e44698ab" impala-user@cloudera.org</span ...
    Darren LoDarren Lo
    Sep 9, 2013 at 3:45 pm
    Sep 11, 2013 at 11:18 am
  • Hi Justin, Our use case is simple in case of Stored Procedures and Cursors. Stored procedures are just being used for running a set of queries together with minimal processing logic and cursors are ...
    Ashish AgrawalAshish Agrawal
    Sep 19, 2013 at 5:45 am
    Sep 16, 2014 at 8:31 am
  • Quick question... What is in the future plans for Impala with regards to the metadata refresh? I personally think it is an inconvenience to manually update it or schedule it on a daily or hourly ...
    Benjamin KimBenjamin Kim
    Sep 23, 2013 at 3:43 pm
    Sep 24, 2013 at 4:38 pm
  • Hi Marcos, Thanks for your question. To ans your question, I strong encourage to visit our Impala-hbase docuemntation in this page ...
    Alan ChoiAlan Choi
    Sep 12, 2013 at 5:23 pm
    Sep 13, 2013 at 5:22 pm
  • Hi Screenthong, We'll reopen the JIRA and look into the issue. But at the mean time, maybe you can try it on Impala 1.1.1? We've fixed quite a few issues since 1.1. Thanks, Alan To unsubscribe from ...
    Alan ChoiAlan Choi
    Sep 11, 2013 at 11:49 pm
    Sep 13, 2013 at 1:44 am
  • Hi, I have encountered strange behaviour when using impala-shell due to invalid query string encoding. For instance: select count(*) from <table where content like '%ąśćó%'; Yields error: Unknown ...
    Radosław SypeńRadosław Sypeń
    Sep 30, 2013 at 12:35 pm
    Sep 30, 2013 at 3:49 pm
  • Hi, The system timezone on our datanodes is UTC. I d like the date and time functions on impala to use the local timezone ie 'America/Los_Angeles'. I added export TZ='America/Los_Angeles' to the ...
    Sam WilliamSam William
    Sep 18, 2013 at 9:56 pm
    Sep 19, 2013 at 6:15 pm
  • Hi, I am busy considering Impala as a serving layer for our stack. A serving layer as described by Nathan Marz's Lambda architecture (http://nathanmarz.com/blog/how-to-beat-the-cap-theorem.html). The ...
    Quinton AndersonQuinton Anderson
    Sep 15, 2013 at 11:17 pm
    Sep 17, 2013 at 6:33 pm
  • 3:26:39.481 PMINFOthrift-util.cc:97 TSocket::read() recv() <Host: hadoop-testing-06 Port: 22000 Connection timed out 3:26:39.482 PMINFOclient-cache.cc:98 CreateClient(): adding new client for ...
    Adline D'SilvaAdline D'Silva
    Sep 6, 2013 at 7:34 am
    Sep 10, 2013 at 1:51 am
  • Hi Eric, For your query, the aggregation nodes are the ones that consumes memory, and the memory consumption depends on the number of distinct group you have. So, ~200MB isn't unreasoanble. The total ...
    Alan ChoiAlan Choi
    Sep 9, 2013 at 7:16 pm
    Sep 10, 2013 at 1:32 am
  • Getting further, after adding all the jars from the parquet github and restarting hiverserver2 and hivemetastore I now get the following when running analyze statistics and a select on a Parquet ...
    Andrew StevensonAndrew Stevenson
    Sep 5, 2013 at 10:37 am
    Sep 6, 2013 at 8:11 am
  • Hi, I plan to benchmark Impala on "Fat Nodes", this is, nodes with 64 cores and 20 disks ( still far from the limit, i.e., in the code we can see a bit-mask that sets the limit to 64 disks). Looking ...
    Romeu KerstenRomeu Kersten
    Sep 14, 2013 at 12:48 am
    Sep 17, 2013 at 11:50 pm
  • Thanks Greg, I'll reorder the select. Does this explain the low network throughput? I ran a iperf test which showed 915Mib/sec. From: Greg Rahn Sent: 03/09/2013 22:58 To: <span ...
    Andrew StevensonAndrew Stevenson
    Sep 4, 2013 at 6:40 am
    Sep 4, 2013 at 8:09 pm
  • Hi, I have Impala 1.0 installed. Can I do LZO compression here? Thanks in advance. Nancy To unsubscribe from this group and stop receiving emails from it, send an email to ...
    Nancy jeanNancy jean
    Sep 3, 2013 at 6:22 am
    Sep 3, 2013 at 8:45 pm
  • Hi, Im trying to get Impala working with out custom setup of CDH4.3.1. Im able to create tables on Impala and query them from both Impala and Hive. However, tables created from Hive dont seem to work ...
    Sam WilliamSam William
    Sep 3, 2013 at 7:43 pm
    Sep 3, 2013 at 7:57 pm
  • Yes, setup a 32bit and a 64bit with the same name (32bit is needed for Visual studio SSDT). The 64bit is used at run time. http://ayadshammout.com/2013/07/13/ssas-tabular-using-hiveodbc-connection/ I ...
    Andrew StevensonAndrew Stevenson
    Sep 2, 2013 at 7:09 am
    Sep 2, 2013 at 8:06 am
  • Hi Impala Users, Please take a look out our blog post today capturing Impala's upcoming roadmap baesed on feedback from customers such as yourselves ...
    Justin EricksonJustin Erickson
    Sep 24, 2013 at 5:35 pm
    Sep 24, 2013 at 5:35 pm
  • Hey guys I am trying to install Impala using cloudera manager. Now everything got install, but at last step it is showing that "Failed to receive heartbeat from agent". In my /etc/hosts/ file ...
    Karam anejaKaram aneja
    Sep 17, 2013 at 2:08 pm
    Sep 17, 2013 at 2:08 pm
  • Hi Ashish, Can you elaborate a more on how you are using stored procedures in the workloads you are trying to bring to Hadoop? For cursors, can you provide details on how you use them, specifically ...
    Justin EricksonJustin Erickson
    Sep 16, 2013 at 8:31 pm
    Sep 16, 2013 at 8:31 pm
  • Hi, Personally I'd be really interested in ROW_NUMBER. LEAD and LAG would be really helpful as well but not as critical. Den torsdagen den 12:e september 2013 kl. 19:38:25 UTC+2 skrev Justin ...
    Fredrik RagnarFredrik Ragnar
    Sep 13, 2013 at 7:50 am
    Sep 13, 2013 at 7:50 am
  • Hi, There are two things you need to check: 1. Make sure that Impalad knows that the list of live backends. Go to http://<impalad host :25000/backends. You should see 2 entires. 2. Make sure that the ...
    Alan ChoiAlan Choi
    Sep 10, 2013 at 8:58 pm
    Sep 10, 2013 at 8:58 pm
  • Each impalad gets the HDFS block metadata from the NameNode and keeps this data cached, hence the need for the ...
    Greg RahnGreg Rahn
    Sep 10, 2013 at 5:16 am
    Sep 10, 2013 at 5:16 am
  • Sounds like impalad might be paging. You should check what else is running on the system and how big the impalad process is at the time that you're experiencing this slowdown. To unsubscribe from ...
    Marcel KornackerMarcel Kornacker
    Sep 10, 2013 at 12:09 am
    Sep 10, 2013 at 12:09 am
  • Dear all, following previous query, a naive question, any chance to roughly estimate the memory usage of particular sql query before actually execute on distributed nodes. normally the query we're ...
    Jason shihJason shih
    Sep 7, 2013 at 12:23 pm
    Sep 7, 2013 at 12:23 pm
  • Here's an example: select * from (select * from T limit 10) u where u.id < 5 In this case, the predicate can't be evaluated by the scan of T (prior to applying the limit), because this would change ...
    Marcel KornackerMarcel Kornacker
    Sep 5, 2013 at 4:00 pm
    Sep 5, 2013 at 4:00 pm
  • It's working now thanks. To unsubscribe from this group and stop receiving emails from it, send an email to impala-user+unsubscribe@cloudera.org.
    Andrew StevensonAndrew Stevenson
    Sep 5, 2013 at 10:38 am
    Sep 5, 2013 at 10:38 am
  • Hi Jeff, The error you are getting is actually an error from the Hive Metastore. I'm not sure what got you into this state, but if you send the Impalad + Hive Metastore logs it would help us to ...
    Lenni KuffLenni Kuff
    Sep 4, 2013 at 4:41 pm
    Sep 4, 2013 at 4:41 pm
  • Thanks. From: Sean Mackrory Sent: 03/09/2013 20:11 To: <span class="m_body_email_addr" title="4f11bca44eddf56fc4f71409e44698ab" impala-user@cloudera.org</span Subject: Re: LZO version error with ...
    Andrew StevensonAndrew Stevenson
    Sep 3, 2013 at 6:34 pm
    Sep 3, 2013 at 6:34 pm
  • Thanks for reporting this, Andrew. We were able to reproduce the issue with that parcel and have re-released it - it appears to be working well now! To unsubscribe from this group and stop receiving ...
    Sean MackrorySean Mackrory
    Sep 3, 2013 at 6:11 pm
    Sep 3, 2013 at 6:11 pm
  • - AverageScannerThreadConcurrency: 4.42 - BytesRead: 7.24 GB - MemoryUsed: 20.25 KB - NumDisksAccessed: 1 - PerReadThreadRawHdfsThroughput: 351.69 MB/sec do you know why the first time to run impalad ...
    IctwangleiIctwanglei
    Sep 3, 2013 at 2:59 am
    Sep 3, 2013 at 2:59 am
  • The hive drivers work with linked servers just not the impala ones. From: <span class="m_body_email_addr" title="cbefed69ffc610cf635eb9c7ae63f35d" chiewyea@gmail.com</span Sent: ‎02/‎09/‎2013 10:06 ...
    Andrew StevensonAndrew Stevenson
    Sep 2, 2013 at 9:20 am
    Sep 2, 2013 at 9:20 am
Group Navigation
period‹ prev | Sep 2013 | next ›
Group Overview
groupimpala-user @
categorieshadoop
discussions39
posts145
users48
websitecloudera.com
irc#hadoop

48 users for September 2013

Alan Choi: 12 posts Karam aneja: 10 posts Marcel Kornacker: 10 posts Justin Erickson: 8 posts Andrew Stevenson: 7 posts Greg Rahn: 7 posts John Russell: 7 posts Lenni Kuff: 6 posts Eric Huang: 5 posts Abhishek Venkatramanan: 4 posts Nishant Patel: 4 posts Sam William: 4 posts Skye Wanderman-Milne: 4 posts Xiaomei song: 4 posts Yu Dongmin: 4 posts Puneethabm: 3 posts Bps cdh: 3 posts X6i4uybz: 2 posts Chiewyea: 2 posts Adline D'Silva: 2 posts
show more