You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@hudi.apache.org by "Selvaraj periyasamy (Jira)" <ji...@apache.org> on 2020/08/28 23:41:00 UTC
[jira] [Commented] (HUDI-1232) Caching takes lot of time
[ https://issues.apache.org/jira/browse/HUDI-1232?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17186849#comment-17186849 ]
Selvaraj periyasamy commented on HUDI-1232:
-------------------------------------------
Thanks Balaji. My hadoop environment is still running with spark 2.3. Can I run 0.6.0 on spark 2.3?
For issue 1: I am able to manage it with spark glob read, instead of hive read. With this approach, I am good with this approach.
Issue 2: I see the performance issue while writing into the COW table. This is purely write and no read involved. Attached the write logs. The more and more my target has partitions, I am noticing a spike in write time. The fix #1919 mentioned is applicable for writing as well.
Thanks,
Selva[^hudiLogs.txt]
> Caching takes lot of time
> -------------------------
>
> Key: HUDI-1232
> URL: https://issues.apache.org/jira/browse/HUDI-1232
> Project: Apache Hudi
> Issue Type: Bug
> Reporter: Selvaraj periyasamy
> Priority: Major
> Attachments: log1.txt
>
>
> I am using Hudi 0.5.0.
>
> Issue 1) I have source 15 different transaction source tables written using COPY_ON_WRITE and all of them are partitioned by transaction_day, transaction_hour. Eat one of them are having more than 6 month data. I need to join all of them and few of the tables would have to looked back up to 6 months for join condition. When I do the join and execute, I see the attached logs rolling for longer time. Say example for one table to list all the files, it takes more than 5 mins for listing down the files before initiating join. and it is happening in sequential flow. Wen I had to join 6 months data from two different tables, it easily takes more than 10 mins to list the file. I have grepped some the lines and attached. File name is log1.txt .
>
> Issue 2) After above joining all 15 tables , am writing then into another huh target table called trr. The other slowness I am seeing is that, as the number of partitions is growing in trr table, I see below logs rolling for all individual partitions even though my write is on only couple of partitions and it takes unto 4 to 5 mins. I pasted only few of them alone. I am wondering , in future , I would ave 3 years worth of data, and write will be very slow every time I write into only couple of partitions.
>
> 20/08/27 02:08:22 INFO HoodieTableConfig: Loading dataset properties from hdfs://oprhqanameservice/projects/cdp/data/cdp_reporting/trr/.hoodie/hoodie.properties
> 20/08/27 02:08:22 INFO HoodieTableMetaClient: Finished Loading Table of type COPY_ON_WRITE from hdfs://oprhqanameservice/projects/cdp/data/cdp_reporting/trr
> 20/08/27 02:08:22 INFO HoodieActiveTimeline: Loaded instants java.util.stream.ReferencePipeline$Head@fed0a8b
> 20/08/27 02:08:22 INFO HoodieTableFileSystemView: Adding file-groups for partition :20200714/01, #FileGroups=1
> 20/08/27 02:08:22 INFO AbstractTableFileSystemView: addFilesToView: NumFiles=4, FileGroupsCreationTime=0, StoreTimeTaken=1
> 20/08/27 02:08:22 INFO HoodieROTablePathFilter: Based on hoodie metadata from base path: hdfs://oprhqanameservice/projects/cdp/data/cdp_reporting/trr, caching 1 files under
> hdfs://oprhqanameservice/projects/cdp/data/cdp_reporting/trr/20200714/01
> 20/08/27 02:08:22 INFO HoodieTableMetaClient: Loading HoodieTableMetaClient from hdfs://oprhqanameservice/projects/cdp/data/cdp_reporting/trr
> 20/08/27 02:08:22 INFO FSUtils: Hadoop Configuration: fs.defaultFS: [hdfs://oprhqanameservice], Config:[Configuration: core-default.xml, core-site.xml, mapred-default.xml, m
> apred-site.xml, yarn-default.xml, yarn-site.xml, hdfs-default.xml, hdfs-site.xml], FileSystem: [DFS[DFSClient[clientName=DFSClient_NONMAPREDUCE_-778362260_1, ugi=svchdc36q@V
> ISA.COM (auth:KERBEROS)]]]
> 20/08/27 02:08:22 INFO HoodieTableConfig: Loading dataset properties from hdfs://oprhqanameservice/projects/cdp/data/cdp_reporting/trr/.hoodie/hoodie.properties
> 20/08/27 02:08:22 INFO HoodieTableMetaClient: Finished Loading Table of type COPY_ON_WRITE from hdfs://oprhqanameservice/projects/cdp/data/cdp_reporting/trr
> 20/08/27 02:08:22 INFO HoodieActiveTimeline: Loaded instants java.util.stream.ReferencePipeline$Head@285c67a9
> 20/08/27 02:08:22 INFO HoodieTableFileSystemView: Adding file-groups for partition :20200714/02, #FileGroups=1
> 20/08/27 02:08:22 INFO AbstractTableFileSystemView: addFilesToView: NumFiles=4, FileGroupsCreationTime=0, StoreTimeTaken=0
> 20/08/27 02:08:22 INFO HoodieROTablePathFilter: Based on hoodie metadata from base path: hdfs://oprhqanameservice/projects/cdp/data/cdp_reporting/trr, caching 1 files under
> hdfs://oprhqanameservice/projects/cdp/data/cdp_reporting/trr/20200714/02
> 20/08/27 02:08:22 INFO HoodieTableMetaClient: Loading HoodieTableMetaClient from hdfs://oprhqanameservice/projects/cdp/data/cdp_reporting/trr
> 20/08/27 02:08:22 INFO FSUtils: Hadoop Configuration: fs.defaultFS: [hdfs://oprhqanameservice], Config:[Configuration: core-default.xml, core-site.xml, mapred-default.xml, m
> apred-site.xml, yarn-default.xml, yarn-site.xml, hdfs-default.xml, hdfs-site.xml], FileSystem: [DFS[DFSClient[clientName=DFSClient_NONMAPREDUCE_-778362260_1, ugi=svchdc36q@V
> ISA.COM (auth:KERBEROS)]]]
> 20/08/27 02:08:22 INFO HoodieTableConfig: Loading dataset properties from hdfs://oprhqanameservice/projects/cdp/data/cdp_reporting/trr/.hoodie/hoodie.properties
> 20/08/27 02:08:22 INFO HoodieTableMetaClient: Finished Loading Table of type COPY_ON_WRITE from hdfs://oprhqanameservice/projects/cdp/data/cdp_reporting/trr
> 20/08/27 02:08:22 INFO HoodieActiveTimeline: Loaded instants java.util.stream.ReferencePipeline$Head@2edd9c8
> 20/08/27 02:08:22 INFO HoodieTableFileSystemView: Adding file-groups for partition :20200714/03, #FileGroups=1
> 20/08/27 02:08:22 INFO AbstractTableFileSystemView: addFilesToView: NumFiles=4, FileGroupsCreationTime=1, StoreTimeTaken=0
> 20/08/27 02:08:22 INFO HoodieROTablePathFilter: Based on hoodie metadata from base path: hdfs://oprhqanameservice/projects/cdp/data/cdp_reporting/trr, caching 1 files under hdfs://oprhqanameservice/projects/cdp/data/cdp_reporting/trr/20200714/03
> 20/08/27 02:08:22 INFO HoodieTableMetaClient: Loading HoodieTableMetaClient from hdfs://oprhqanameservice/projects/cdp/data/cdp_reporting/trr
> 20/08/27 02:08:22 INFO FSUtils: Hadoop Configuration: fs.defaultFS: [hdfs://oprhqanameservice], Config:[Configuration: core-default.xml, core-site.xml, mapred-default.xml, mapred-site.xml, yarn-default.xml, yarn-site.xml, hdfs-default.xml, hdfs-site.xml], FileSystem: [DFS[DFSClient[clientName=DFSClient_NONMAPREDUCE_-778362260_1, ugi=svchdc36q@VISA.COM (auth:KERBEROS)]]]
>
>
> could someone provide more insight on ow to make these things work faster and make it scalable when the number of partitions is increasing?
>
> I have below setting in stark session.
>
> sparkSession.sparkContext.hadoopConfiguration.setClass("mapreduce.input.pathFilter.class", classOf[org.apache.hudi.hadoop.HoodieROTablePathFilter], classOf[org.apache.hadoop.fs.PathFilter]);
>
> and also using --conf spark.sql.hive.convertMetastoreParquet=false in spark2-submit
>
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)