Created on 11-21-2014 10:48 PM - edited 09-16-2022 02:14 AM
Hi, I have 2 node (Namenode + 2 Datanode) setup for Hadoop (CDH 5.2) , my servers are centos 6.5 64 bit (2 core with 4GB RAM). I also have one more cluster (CDH 4.2) having similar configuration.
Now my problem is that, I have one sql query which is running flawless on CDH 4.2 cluster but giving "Backend 6:Memory Limit Exceeded" in new CDH 5.2 cluster.
I have also increased datanode RAM upto 8GB and set MEM_LIMIT=6G in impala, but still getting same error.
I have also tried "explain" statement to check memory usage by query and it is showing per node memory 2.3G, then why I am getting this error.
Created 12-18-2014 10:24 AM
Hi Nagendra,
I replied to you over email. We are trying to understand and fix the 'Memory Limit Exceeded' errors. Can you please disable the partitioned_hash_join and partitioned_aggregation, by restarting Impala with those options "--enable_partitioned_aggregation=false --enable_partitioned_hash_join=false" and let me know if the queries run successfully?
Regards,
-Ippokratis.
Created 11-22-2014 09:19 AM
Hi,
Is the query that is failing running in isolation or are other workloads running at the same time? Can you post the profile information for that query and any relevant sections from the impalad log?
Dimitris
Created 11-24-2014 03:30 AM
Hi Dimitris,
This query is running in isolation, there are no other workload. below are the query profile and impala logs.
[server2.mydomain.com:21000] > profile;
Query Runtime Profile:
Query (id=92476f3f27f301ac:a1a07af98ab08da5):
Summary:
Session ID: 2d4ede846bdcd77d:c587f0a88636d6be
Session Type: BEESWAX
Start Time: 2014-11-24 16:48:47.207538000
End Time: 2014-11-24 16:49:02.334840000
Query Type: QUERY
Query State: EXCEPTION
Query Status: Memory limit exceeded
Cannot perform hash aggregation. Partitioned input data too many times. This could mean there is too much skew in the data or the memory limit is set too low.
Impala Version: impalad version 2.0.0-cdh5 RELEASE (build ecf30af0b4d6e56ea80297df2189367ada6b7da7)
User: root
Connected User: root
Delegated User:
Network Address: ::ffff:10.6.102.133:59128
Default Db: default
Sql Statement: select case when d.ip_location is null then 'internet' else d.ip_location end, c.access_date, c.usr_id from MST_LOCATION d right outer join (select distinct ip_address, a.access_date, a.usr_id,a.cnt from logs_sc1 a inner join (select access_date,a.usr_id from (select a.ip_address,case when a.ip_location is null then 'internet' else a.ip_location end ,access_date,b.usr_id,b.cnt FROM logs_sc1 B Left outer join MST_LOCATION A on SUBSTR(a.ip_address,1,5) = SUBSTR(b.ip_address,1,5) WHERE b.ip_address <> "10.6.1.230" and b.sp_used is not null and page_used like '%welcome.jsp%' group by a.ip_address, a.ip_location, access_date, b.usr_id,b.cnt) A group by access_date, a.usr_id, cnt having count(1) > 1)b on a.access_date =b.access_date and a.usr_id=b.usr_id )c on SUBSTR(c.ip_address,1,5) = SUBSTR(d.ip_address,1,5) group by d.ip_location, c.access_date , c.usr_id order by c.usr_id,c.access_date limit 10000
Coordinator: server2:22000
Plan:
----------------
Estimated Per-Host Requirements: Memory=2.59GB VCores=2
WARNING: The following tables are missing relevant table and/or column statistics.
default.logs_sc1, default.mst_location
F09:PLAN FRAGMENT [UNPARTITIONED]
24:MERGING-EXCHANGE [UNPARTITIONED]
order by: c.usr_id ASC, c.access_date ASC
limit: 10000
hosts=2 per-host-mem=unavailable
tuple-ids=11 row-size=45B cardinality=10000
F08:PLAN FRAGMENT [HASH(d.ip_location,c.access_date,c.usr_id)]
DATASTREAM SINK [FRAGMENT=F09, EXCHANGE=24, UNPARTITIONED]
11:TOP-N [LIMIT=10000]
| order by: c.usr_id ASC, c.access_date ASC
| hosts=2 per-host-mem=439.45KB
| tuple-ids=11 row-size=45B cardinality=10000
|
23:AGGREGATE [FINALIZE]
| group by: d.ip_location, c.access_date, c.usr_id
| hosts=2 per-host-mem=128.00MB
| tuple-ids=10 row-size=45B cardinality=unavailable
|
22:EXCHANGE [HASH(d.ip_location,c.access_date,c.usr_id)]
hosts=2 per-host-mem=0B
tuple-ids=10 row-size=45B cardinality=unavailable
F07:PLAN FRAGMENT [HASH(substr(d.ip_address, 1, 5))]
DATASTREAM SINK [FRAGMENT=F08, EXCHANGE=22, HASH(d.ip_location,c.access_date,c.usr_id)]
10:AGGREGATE
| group by: d.ip_location, a.access_date, a.usr_id
| hosts=2 per-host-mem=128.00MB
| tuple-ids=10 row-size=45B cardinality=unavailable
|
09:HASH JOIN [RIGHT OUTER JOIN, PARTITIONED]
| hash predicates: substr(d.ip_address, 1, 5) = substr(ip_address, 1, 5)
| hosts=2 per-host-mem=2.00GB
| tuple-ids=0N,8 row-size=77B cardinality=unavailable
|
|--21:EXCHANGE [HASH(substr(ip_address, 1, 5))]
| hosts=2 per-host-mem=0B
| tuple-ids=8 row-size=47B cardinality=unavailable
|
20:EXCHANGE [HASH(substr(d.ip_address, 1, 5))]
hosts=2 per-host-mem=0B
tuple-ids=0 row-size=30B cardinality=unavailable
F06:PLAN FRAGMENT [HASH(ip_address,a.access_date,a.usr_id,a.cnt)]
DATASTREAM SINK [FRAGMENT=F07, EXCHANGE=21, HASH(substr(ip_address, 1, 5))]
19:AGGREGATE [FINALIZE]
| group by: ip_address, a.access_date, a.usr_id, a.cnt
| hosts=2 per-host-mem=128.00MB
| tuple-ids=8 row-size=47B cardinality=unavailable
|
18:EXCHANGE [HASH(ip_address,a.access_date,a.usr_id,a.cnt)]
hosts=2 per-host-mem=0B
tuple-ids=8 row-size=47B cardinality=unavailable
F01:PLAN FRAGMENT [RANDOM]
DATASTREAM SINK [FRAGMENT=F06, EXCHANGE=18, HASH(ip_address,a.access_date,a.usr_id,a.cnt)]
08:AGGREGATE
| group by: ip_address, a.access_date, a.usr_id, a.cnt
| hosts=2 per-host-mem=128.00MB
| tuple-ids=8 row-size=47B cardinality=unavailable
|
07:HASH JOIN [INNER JOIN, BROADCAST]
| hash predicates: a.access_date = access_date, a.usr_id = a.usr_id
| hosts=2 per-host-mem=2.00GB
| tuple-ids=1,6 row-size=87B cardinality=unavailable
|
|--17:EXCHANGE [BROADCAST]
| hosts=2 per-host-mem=0B
| tuple-ids=6 row-size=40B cardinality=unavailable
|
01:SCAN HDFS [default.logs_sc1 a, RANDOM]
partitions=1/1 size=968.40MB
table stats: unavailable
column stats: unavailable
hosts=2 per-host-mem=352.00MB
tuple-ids=1 row-size=47B cardinality=unavailable
F05:PLAN FRAGMENT [HASH(access_date,a.usr_id,cnt)]
DATASTREAM SINK [FRAGMENT=F01, EXCHANGE=17, BROADCAST]
16:AGGREGATE [FINALIZE]
| output: count:merge(1)
| group by: access_date, a.usr_id, cnt
| having: count(1) > 1
| hosts=2 per-host-mem=128.00MB
| tuple-ids=6 row-size=40B cardinality=unavailable
|
15:EXCHANGE [HASH(access_date,a.usr_id,cnt)]
hosts=2 per-host-mem=0B
tuple-ids=6 row-size=40B cardinality=unavailable
F04:PLAN FRAGMENT [HASH(a.ip_address,a.ip_location,access_date,b.usr_id,b.cnt)]
DATASTREAM SINK [FRAGMENT=F05, EXCHANGE=15, HASH(access_date,a.usr_id,cnt)]
06:AGGREGATE
| output: count(1)
| group by: access_date, b.usr_id, b.cnt
| hosts=2 per-host-mem=128.00MB
| tuple-ids=6 row-size=40B cardinality=unavailable
|
14:AGGREGATE [FINALIZE]
| group by: a.ip_address, a.ip_location, access_date, b.usr_id, b.cnt
| hosts=2 per-host-mem=128.00MB
| tuple-ids=4 row-size=62B cardinality=unavailable
|
13:EXCHANGE [HASH(a.ip_address,a.ip_location,access_date,b.usr_id,b.cnt)]
hosts=2 per-host-mem=0B
tuple-ids=4 row-size=62B cardinality=unavailable
F02:PLAN FRAGMENT [RANDOM]
DATASTREAM SINK [FRAGMENT=F04, EXCHANGE=13, HASH(a.ip_address,a.ip_location,access_date,b.usr_id,b.cnt)]
05:AGGREGATE
| group by: a.ip_address, a.ip_location, access_date, b.usr_id, b.cnt
| hosts=2 per-host-mem=128.00MB
| tuple-ids=4 row-size=62B cardinality=unavailable
|
04:HASH JOIN [LEFT OUTER JOIN, BROADCAST]
| hash predicates: substr(b.ip_address, 1, 5) = substr(a.ip_address, 1, 5)
| hosts=2 per-host-mem=2.00GB
| tuple-ids=2,3N row-size=107B cardinality=unavailable
|
|--12:EXCHANGE [BROADCAST]
| hosts=2 per-host-mem=0B
| tuple-ids=3 row-size=30B cardinality=unavailable
|
02:SCAN HDFS [default.logs_sc1 B, RANDOM]
partitions=1/1 size=968.40MB
predicates: b.ip_address != '10.6.1.230', b.sp_used IS NOT NULL, page_used LIKE '%welcome.jsp%'
table stats: unavailable
column stats: unavailable
hosts=2 per-host-mem=480.00MB
tuple-ids=2 row-size=77B cardinality=unavailable
F03:PLAN FRAGMENT [RANDOM]
DATASTREAM SINK [FRAGMENT=F02, EXCHANGE=12, BROADCAST]
03:SCAN HDFS [default.mst_location A, RANDOM]
partitions=1/1 size=250B
table stats: unavailable
column stats: unavailable
hosts=2 per-host-mem=32.00MB
tuple-ids=3 row-size=30B cardinality=unavailable
F00:PLAN FRAGMENT [RANDOM]
DATASTREAM SINK [FRAGMENT=F07, EXCHANGE=20, HASH(substr(d.ip_address, 1, 5))]
00:SCAN HDFS [default.mst_location d, RANDOM]
partitions=1/1 size=250B
table stats: unavailable
column stats: unavailable
hosts=2 per-host-mem=32.00MB
tuple-ids=0 row-size=30B cardinality=unavailable
----------------
Estimated Per-Host Mem: 2785017856
Estimated Per-Host VCores: 2
Tables Missing Stats: default.logs_sc1,default.mst_location
Admission result: Admitted immediately
Request Pool: default-pool
ExecSummary:
Operator #Hosts Avg Time Max Time #Rows Est. #Rows Peak Mem Est. Peak Mem Detail
------------------------------------------------------------------------------------------------------------------------------------
24:MERGING-EXCHANGE 1 29.64us 29.64us 0 10.00K 0 -1.00 B UNPARTITIONED
11:TOP-N 1 16.211us 16.211us 0 10.00K 4.00 KB 439.45 KB
23:AGGREGATE 1 248.979ms 248.979ms 0 -1 6.25 MB 128.00 MB FINALIZE
22:EXCHANGE 1 0ns 0ns 0 -1 0 0 HASH(d.ip_location,c.access...
10:AGGREGATE 1 251.322ms 251.322ms 0 -1 6.25 MB 128.00 MB
09:HASH JOIN 1 0ns 0ns 0 -1 8.00 KB 2.00 GB RIGHT OUTER JOIN, PARTITIONED
|--21:EXCHANGE 1 0ns 0ns 0 -1 0 0 HASH(substr(ip_address, 1, 5))
| 19:AGGREGATE 2 229.878ms 252.185ms 0 -1 6.25 MB 128.00 MB FINALIZE
| 18:EXCHANGE 2 0ns 0ns 0 -1 0 0 HASH(ip_address,a.access_da...
| 08:AGGREGATE 2 244.661ms 252.171ms 0 -1 6.25 MB 128.00 MB
| 07:HASH JOIN 2 5.549ms 6.881ms 0 -1 8.00 KB 2.00 GB INNER JOIN, BROADCAST
| |--17:EXCHANGE 2 0ns 0ns 0 -1 0 0 BROADCAST
| | 16:AGGREGATE 2 220.740ms 288.538ms 0 -1 6.25 MB 128.00 MB FINALIZE
| | 15:EXCHANGE 2 0ns 0ns 0 -1 0 0 HASH(access_date,a.usr_id,cnt)
| | 06:AGGREGATE 2 204.343ms 254.801ms 0 -1 6.25 MB 128.00 MB
| | 14:AGGREGATE 2 3.274ms 3.331ms 0 -1 6.25 MB 128.00 MB FINALIZE
| | 13:EXCHANGE 2 0ns 0ns 0 -1 0 0 HASH(a.ip_address,a.ip_loca...
| | 05:AGGREGATE 2 0ns 0ns 0 -1 569.32 MB 128.00 MB
| | 04:HASH JOIN 2 134.457ms 215.946ms 904.78K -1 21.29 MB 2.00 GB LEFT OUTER JOIN, BROADCAST
| | |--12:EXCHANGE 2 12.323us 13.368us 11 -1 0 0 BROADCAST
| | | 03:SCAN HDFS 1 251.71ms 251.71ms 11 -1 54.00 KB 32.00 MB default.mst_location A
| | 02:SCAN HDFS 2 7s192ms 7s231ms 904.78K -1 156.00 MB 480.00 MB default.logs_sc1 B
| 01:SCAN HDFS 2 115.86us 133.579us 0 -1 0 352.00 MB default.logs_sc1 a
20:EXCHANGE 1 176.181us 176.181us 0 -1 0 0 HASH(substr(d.ip_address, 1...
00:SCAN HDFS 1 282.55ms 282.55ms 11 -1 54.00 KB 32.00 MB default.mst_location d
Query Timeline: 15s135ms
- Start execution: 63.453us (63.453us)
- Planning finished: 161.169ms (161.105ms)
- Submit for admission: 161.444ms (274.910us)
- Completed admission: 161.454ms (10.504us)
- Ready to start remote fragments: 164.93ms (2.639ms)
- Remote fragments started: 2s575ms (2s411ms)
- Rows available: 14s726ms (12s150ms)
- Unregister query: 15s127ms (401.62ms)
Impala Log:
==> /var/log/impala/impalad.INFO <==
W1124 16:49:29.641473 26908 HdfsScanNode.java:720] Per-host mem cost 528.00MB exceeded per-host upper bound 480.00MB.
==> /var/log/impala/impalad.server2.impala.log.INFO.20141124-155700.26843 <==
W1124 16:49:29.641473 26908 HdfsScanNode.java:720] Per-host mem cost 528.00MB exceeded per-host upper bound 480.00MB.
Regards,
Nagendra Rai
Created 11-24-2014 11:26 AM
Hi,
Quick question. Do you have statistics in the tables you're querying? It doesn't look like it. Can you run compute stats and see if the problem persists. Also, the log messages you posted are just warnings. The problem is not with the scan nodes but with the computation of the hash join.
Dimitris
Created 11-24-2014 12:02 PM
Hi Dimitris,
I am using same table with same query in hive and getting the ouput. I will try compute stat and will let you know.
Thanks & Regards,
Nagendra Rai
Created 11-25-2014 03:00 AM
Hi Dimitris,
After running "compute stat logs_sc1" I was getting below mentioned error for all queries having "logs_sc1" table in it.
ERROR: AnalysisException: Failed to load metadata for table: default.logs_sc1
CAUSED BY: TableLoadingException: Failed to load metadata for table: logs_sc1
CAUSED BY: TTransportException: java.net.SocketException: Broken pipe
CAUSED BY: SocketException: Broken pipe
Then I set below mentioned property in hive-site.xml and restarted hive metastore to remove this error.
<property>
<name>hive.metastore.try.direct.sql</name>
<value>false</value>
</property>
But I was still getting "Memory Limit Exceeded" error, then I executed explain statement and was surprised to see that Estimated Per-Host Requirements: Memory becomes 14.43GB from 2.83GB.
Finally I was able to run this query by increasing impala node (server) memory up to 16GB and setting impala memory to 15GB "set MEM_LIMIT=15G;"
Now my query is why CDH 5.2 and impala 2 require that much memory for same data and same query which is running smooth on CDH 4.2 and impala 1.1.1 server having 4GB memory
Explain statement output from CDH 5 for table with compute stat:
Query: explain select case when d.ip_location is null then 'internet' else d.ip_location end, c.access_date, c.usr_id from MST_LOCATION d right outer join (select distinct ip_address, a.access_date, a.usr_id,a.cnt from logs_sc1 a inner join (select access_date,a.usr_id from (select a.ip_address,case when a.ip_location is null then 'internet' else a.ip_location end ,access_date,b.usr_id,b.cnt FROM logs_sc1 B Left outer join MST_LOCATION A on SUBSTR(a.ip_address,1,5) = SUBSTR(b.ip_address,1,5) WHERE b.ip_address <> "10.6.1.230" and b.sp_used is not null and page_used like '%welcome.jsp%' group by a.ip_address, a.ip_location, access_date, b.usr_id,b.cnt) A group by access_date, a.usr_id, cnt having count(1) > 1)b on a.access_date =b.access_date and a.usr_id=b.usr_id )c on SUBSTR(c.ip_address,1,5) = SUBSTR(d.ip_address,1,5) group by d.ip_location, c.access_date , c.usr_id order by c.usr_id,c.access_date limit 10000 +-------------------------------------------------------------------------------------------------------+ | Explain String | +-------------------------------------------------------------------------------------------------------+ | Estimated Per-Host Requirements: Memory=14.43GB VCores=2 | | WARNING: The following tables are missing relevant table and/or column statistics. | | default.mst_location | | | | 23:MERGING-EXCHANGE [UNPARTITIONED] | | | order by: c.usr_id ASC, c.access_date ASC | | | limit: 10000 | | | | | 11:TOP-N [LIMIT=10000] | | | order by: c.usr_id ASC, c.access_date ASC | | | | | 22:AGGREGATE [FINALIZE] | | | group by: d.ip_location, c.access_date, c.usr_id | | | | | 21:EXCHANGE [HASH(d.ip_location,c.access_date,c.usr_id)] | | | | | 10:AGGREGATE | | | group by: d.ip_location, a.access_date, a.usr_id | | | | | 09:HASH JOIN [LEFT OUTER JOIN, BROADCAST] | | | hash predicates: substr(ip_address, 1, 5) = substr(d.ip_address, 1, 5) | | | | | |--20:EXCHANGE [BROADCAST] | | | | | | | 00:SCAN HDFS [default.mst_location d] | | | partitions=1/1 size=250B | | | | | 19:AGGREGATE [FINALIZE] | | | group by: ip_address, a.access_date, a.usr_id, a.cnt | | | | | 18:EXCHANGE [HASH(ip_address,a.access_date,a.usr_id,a.cnt)] | | | | | 08:AGGREGATE | | | group by: ip_address, a.access_date, a.usr_id, a.cnt | | | | | 07:HASH JOIN [INNER JOIN, BROADCAST] | | | hash predicates: a.access_date = access_date, a.usr_id = a.usr_id | | | | | |--17:EXCHANGE [BROADCAST] | | | | | | | 16:AGGREGATE [FINALIZE] | | | | output: count:merge(1) | | | | group by: access_date, a.usr_id, cnt | | | | having: count(1) > 1 | | | | | | | 15:EXCHANGE [HASH(access_date,a.usr_id,cnt)] | | | | | | | 06:AGGREGATE | | | | output: count(1) | | | | group by: access_date, b.usr_id, b.cnt | | | | | | | 14:AGGREGATE [FINALIZE] | | | | group by: a.ip_address, a.ip_location, access_date, b.usr_id, b.cnt | | | | | | | 13:EXCHANGE [HASH(a.ip_address,a.ip_location,access_date,b.usr_id,b.cnt)] | | | | | | | 05:AGGREGATE | | | | group by: a.ip_address, a.ip_location, access_date, b.usr_id, b.cnt | | | | | | | 04:HASH JOIN [LEFT OUTER JOIN, BROADCAST] | | | | hash predicates: substr(b.ip_address, 1, 5) = substr(a.ip_address, 1, 5) | | | | | | | |--12:EXCHANGE [BROADCAST] | | | | | | | | | 03:SCAN HDFS [default.mst_location A] | | | | partitions=1/1 size=250B | | | | | | | 02:SCAN HDFS [default.logs_sc1 B] | | | partitions=1/1 size=968.40MB | | | predicates: b.ip_address != '10.6.1.230', b.sp_used IS NOT NULL, page_used LIKE '%welcome.jsp%' | | | | | 01:SCAN HDFS [default.logs_sc1 a] | | partitions=1/1 size=968.40MB | +-------------------------------------------------------------------------------------------------------+ Fetched 73 row(s) in 0.23s
Explain statement output from CDH 5 for table without compute stat (This is different table with same config and data), earlier I was geeting same memory error for this table too, but it is running fine without compute stat statement.
+----------------------------------------------------------------------------------------------------------+ | Explain String | +----------------------------------------------------------------------------------------------------------+ | Estimated Per-Host Requirements: Memory=2.59GB VCores=2 | | WARNING: The following tables are missing relevant table and/or column statistics. | | default.logs_sc, default.mst_location | | | | 24:MERGING-EXCHANGE [UNPARTITIONED] | | | order by: c.usr_id ASC, c.access_date ASC | | | limit: 10000 | | | | | 11:TOP-N [LIMIT=10000] | | | order by: c.usr_id ASC, c.access_date ASC | | | | | 23:AGGREGATE [FINALIZE] | | | group by: d.ip_location, c.access_date, c.usr_id | | | | | 22:EXCHANGE [HASH(d.ip_location,c.access_date,c.usr_id)] | | | | | 10:AGGREGATE | | | group by: d.ip_location, a.access_date, a.usr_id | | | | | 09:HASH JOIN [RIGHT OUTER JOIN, PARTITIONED] | | | hash predicates: substr(d.ip_address, 1, 5) = substr(ip_address, 1, 5) | | | | | |--21:EXCHANGE [HASH(substr(ip_address, 1, 5))] | | | | | | | 19:AGGREGATE [FINALIZE] | | | | group by: ip_address, a.access_date, a.usr_id, a.cnt | | | | | | | 18:EXCHANGE [HASH(ip_address,a.access_date,a.usr_id,a.cnt)] | | | | | | | 08:AGGREGATE | | | | group by: ip_address, a.access_date, a.usr_id, a.cnt | | | | | | | 07:HASH JOIN [INNER JOIN, BROADCAST] | | | | hash predicates: a.access_date = access_date, a.usr_id = a.usr_id | | | | | | | |--17:EXCHANGE [BROADCAST] | | | | | | | | | 16:AGGREGATE [FINALIZE] | | | | | output: count:merge(1) | | | | | group by: access_date, a.usr_id, cnt | | | | | having: count(1) > 1 | | | | | | | | | 15:EXCHANGE [HASH(access_date,a.usr_id,cnt)] | | | | | | | | | 06:AGGREGATE | | | | | output: count(1) | | | | | group by: access_date, b.usr_id, b.cnt | | | | | | | | | 14:AGGREGATE [FINALIZE] | | | | | group by: a.ip_address, a.ip_location, access_date, b.usr_id, b.cnt | | | | | | | | | 13:EXCHANGE [HASH(a.ip_address,a.ip_location,access_date,b.usr_id,b.cnt)] | | | | | | | | | 05:AGGREGATE | | | | | group by: a.ip_address, a.ip_location, access_date, b.usr_id, b.cnt | | | | | | | | | 04:HASH JOIN [LEFT OUTER JOIN, BROADCAST] | | | | | hash predicates: substr(b.ip_address, 1, 5) = substr(a.ip_address, 1, 5) | | | | | | | | | |--12:EXCHANGE [BROADCAST] | | | | | | | | | | | 03:SCAN HDFS [default.mst_location A] | | | | | partitions=1/1 size=250B | | | | | | | | | 02:SCAN HDFS [default.logs_sc B] | | | | partitions=1/1 size=1.12GB | | | | predicates: b.ip_address != '10.6.1.230', b.sp_used IS NOT NULL, page_used LIKE '%welcome.jsp%' | | | | | | | 01:SCAN HDFS [default.logs_sc a] | | | partitions=1/1 size=1.12GB | | | | | 20:EXCHANGE [HASH(substr(d.ip_address, 1, 5))] | | | | | 00:SCAN HDFS [default.mst_location d] | | partitions=1/1 size=250B | +----------------------------------------------------------------------------------------------------------+ Fetched 75 row(s) in 0.26s
Conclusion:
1) Without compute stat output of explain statement was not correct.
2) After compute stat I was able to know real memory requirement.
3) Still confused why CDH5.2 and impala 2 require 4 to 5 times more memory than CDH 4.2 and impala 1.1.1
Regards,
Nagendra Rai
Created 11-25-2014 04:55 AM
Hi Dimitris,
One more update, there seems to be an issue/bug with impala.
Now I am running single impala node with 16GB ram, as mentioned in previous post. But during execution of query my server memory utilization never goes beyond 4.6GB (with cache and buffer) and 2.3GB (without cache and buffer), then why it is asking for 14GB ram without even using it and crashing with “memory exceeded error” if not provided.
Below are server memory utilization details during query execution in impala.
*********** Tue Nov 25 18:06:05 IST 2014 ************* total used free shared buffers cached Mem: 15759 4589 11170 0 24 2303 -/+ buffers/cache: 2261 13498 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:06 IST 2014 ************* total used free shared buffers cached Mem: 15759 4579 11180 0 24 2295 -/+ buffers/cache: 2259 13500 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:07 IST 2014 ************* total used free shared buffers cached Mem: 15759 4580 11179 0 24 2295 -/+ buffers/cache: 2260 13499 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:08 IST 2014 ************* total used free shared buffers cached Mem: 15759 4583 11176 0 24 2295 -/+ buffers/cache: 2262 13496 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:09 IST 2014 ************* total used free shared buffers cached Mem: 15759 4583 11175 0 24 2295 -/+ buffers/cache: 2263 13496 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:10 IST 2014 ************* total used free shared buffers cached Mem: 15759 4583 11175 0 24 2295 -/+ buffers/cache: 2263 13495 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:11 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11173 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:12 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11173 0 24 2295 -/+ buffers/cache: 2265 13493 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:13 IST 2014 ************* total used free shared buffers cached Mem: 15759 4586 11173 0 24 2295 -/+ buffers/cache: 2266 13493 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:14 IST 2014 ************* total used free shared buffers cached Mem: 15759 4586 11173 0 24 2295 -/+ buffers/cache: 2266 13493 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:15 IST 2014 ************* total used free shared buffers cached Mem: 15759 4586 11173 0 24 2295 -/+ buffers/cache: 2266 13493 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:16 IST 2014 ************* total used free shared buffers cached Mem: 15759 4586 11172 0 24 2295 -/+ buffers/cache: 2266 13493 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:17 IST 2014 ************* total used free shared buffers cached Mem: 15759 4587 11172 0 24 2295 -/+ buffers/cache: 2266 13492 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:18 IST 2014 ************* total used free shared buffers cached Mem: 15759 4587 11172 0 24 2295 -/+ buffers/cache: 2266 13492 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:19 IST 2014 ************* total used free shared buffers cached Mem: 15759 4583 11176 0 24 2295 -/+ buffers/cache: 2263 13496 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:20 IST 2014 ************* total used free shared buffers cached Mem: 15759 4583 11176 0 24 2295 -/+ buffers/cache: 2263 13496 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:21 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11174 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:22 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11174 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:23 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11174 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:24 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11173 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:25 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11173 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:26 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11173 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:27 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11173 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:28 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11173 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:29 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11173 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:30 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11173 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:31 IST 2014 ************* total used free shared buffers cached Mem: 15759 4587 11172 0 24 2295 -/+ buffers/cache: 2267 13492 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:32 IST 2014 ************* total used free shared buffers cached Mem: 15759 4587 11172 0 24 2295 -/+ buffers/cache: 2267 13492 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:33 IST 2014 ************* total used free shared buffers cached Mem: 15759 4587 11172 0 24 2295 -/+ buffers/cache: 2267 13492 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:34 IST 2014 ************* total used free shared buffers cached Mem: 15759 4587 11172 0 24 2295 -/+ buffers/cache: 2267 13492 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:35 IST 2014 ************* total used free shared buffers cached Mem: 15759 4587 11172 0 24 2295 -/+ buffers/cache: 2267 13492 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:36 IST 2014 ************* total used free shared buffers cached Mem: 15759 4591 11168 0 24 2302 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:37 IST 2014 ************* total used free shared buffers cached Mem: 15759 4602 11156 0 24 2303 -/+ buffers/cache: 2274 13485 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:38 IST 2014 ************* total used free shared buffers cached Mem: 15759 4603 11156 0 24 2303 -/+ buffers/cache: 2274 13484 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:39 IST 2014 ************* total used free shared buffers cached Mem: 15759 4603 11156 0 24 2303 -/+ buffers/cache: 2274 13484 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:40 IST 2014 ************* total used free shared buffers cached Mem: 15759 4603 11156 0 24 2303 -/+ buffers/cache: 2274 13484 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:41 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2276 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:42 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2276 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:43 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2275 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:44 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2276 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:45 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2276 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:46 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2276 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:47 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2276 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:48 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2276 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:49 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2276 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:50 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2276 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:51 IST 2014 ************* total used free shared buffers cached Mem: 15759 4606 11153 0 24 2303 -/+ buffers/cache: 2277 13482 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:52 IST 2014 ************* total used free shared buffers cached Mem: 15759 4605 11153 0 24 2303 -/+ buffers/cache: 2277 13482 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:53 IST 2014 ************* total used free shared buffers cached Mem: 15759 4605 11153 0 24 2303 -/+ buffers/cache: 2277 13482 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:54 IST 2014 ************* total used free shared buffers cached Mem: 15759 4605 11153 0 24 2303 -/+ buffers/cache: 2277 13482 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:55 IST 2014 ************* total used free shared buffers cached Mem: 15759 4605 11154 0 24 2303 -/+ buffers/cache: 2277 13482 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:56 IST 2014 ************* total used free shared buffers cached Mem: 15759 4605 11154 0 24 2303 -/+ buffers/cache: 2277 13482 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:57 IST 2014 ************* ^C [root@server2 ~]#
Regards,
Nagendra Rai
Created 11-25-2014 05:04 AM
Hi Dimitris,
One more update, there seems to be an issue/bug with impala.
Now I am running single impala node with 16GB ram, as mentioned in previous post. But during execution of query my server memory utilization never goes beyond 4.6GB (with cache and buffer) and 2.3GB (without cache and buffer), then why it is asking for 14GB ram without even using it and crashing with “memory exceeded error” if not provided.
*********** Tue Nov 25 18:06:05 IST 2014 ************* total used free shared buffers cached Mem: 15759 4589 11170 0 24 2303 -/+ buffers/cache: 2261 13498 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:06 IST 2014 ************* total used free shared buffers cached Mem: 15759 4579 11180 0 24 2295 -/+ buffers/cache: 2259 13500 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:07 IST 2014 ************* total used free shared buffers cached Mem: 15759 4580 11179 0 24 2295 -/+ buffers/cache: 2260 13499 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:08 IST 2014 ************* total used free shared buffers cached Mem: 15759 4583 11176 0 24 2295 -/+ buffers/cache: 2262 13496 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:09 IST 2014 ************* total used free shared buffers cached Mem: 15759 4583 11175 0 24 2295 -/+ buffers/cache: 2263 13496 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:10 IST 2014 ************* total used free shared buffers cached Mem: 15759 4583 11175 0 24 2295 -/+ buffers/cache: 2263 13495 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:11 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11173 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:12 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11173 0 24 2295 -/+ buffers/cache: 2265 13493 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:13 IST 2014 ************* total used free shared buffers cached Mem: 15759 4586 11173 0 24 2295 -/+ buffers/cache: 2266 13493 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:14 IST 2014 ************* total used free shared buffers cached Mem: 15759 4586 11173 0 24 2295 -/+ buffers/cache: 2266 13493 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:15 IST 2014 ************* total used free shared buffers cached Mem: 15759 4586 11173 0 24 2295 -/+ buffers/cache: 2266 13493 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:16 IST 2014 ************* total used free shared buffers cached Mem: 15759 4586 11172 0 24 2295 -/+ buffers/cache: 2266 13493 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:17 IST 2014 ************* total used free shared buffers cached Mem: 15759 4587 11172 0 24 2295 -/+ buffers/cache: 2266 13492 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:18 IST 2014 ************* total used free shared buffers cached Mem: 15759 4587 11172 0 24 2295 -/+ buffers/cache: 2266 13492 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:19 IST 2014 ************* total used free shared buffers cached Mem: 15759 4583 11176 0 24 2295 -/+ buffers/cache: 2263 13496 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:20 IST 2014 ************* total used free shared buffers cached Mem: 15759 4583 11176 0 24 2295 -/+ buffers/cache: 2263 13496 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:21 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11174 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:22 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11174 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:23 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11174 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:24 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11173 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:25 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11173 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:26 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11173 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:27 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11173 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:28 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11173 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:29 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11173 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:30 IST 2014 ************* total used free shared buffers cached Mem: 15759 4585 11173 0 24 2295 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:31 IST 2014 ************* total used free shared buffers cached Mem: 15759 4587 11172 0 24 2295 -/+ buffers/cache: 2267 13492 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:32 IST 2014 ************* total used free shared buffers cached Mem: 15759 4587 11172 0 24 2295 -/+ buffers/cache: 2267 13492 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:33 IST 2014 ************* total used free shared buffers cached Mem: 15759 4587 11172 0 24 2295 -/+ buffers/cache: 2267 13492 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:34 IST 2014 ************* total used free shared buffers cached Mem: 15759 4587 11172 0 24 2295 -/+ buffers/cache: 2267 13492 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:35 IST 2014 ************* total used free shared buffers cached Mem: 15759 4587 11172 0 24 2295 -/+ buffers/cache: 2267 13492 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:36 IST 2014 ************* total used free shared buffers cached Mem: 15759 4591 11168 0 24 2302 -/+ buffers/cache: 2265 13494 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:37 IST 2014 ************* total used free shared buffers cached Mem: 15759 4602 11156 0 24 2303 -/+ buffers/cache: 2274 13485 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:38 IST 2014 ************* total used free shared buffers cached Mem: 15759 4603 11156 0 24 2303 -/+ buffers/cache: 2274 13484 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:39 IST 2014 ************* total used free shared buffers cached Mem: 15759 4603 11156 0 24 2303 -/+ buffers/cache: 2274 13484 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:40 IST 2014 ************* total used free shared buffers cached Mem: 15759 4603 11156 0 24 2303 -/+ buffers/cache: 2274 13484 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:41 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2276 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:42 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2276 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:43 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2275 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:44 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2276 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:45 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2276 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:46 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2276 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:47 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2276 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:48 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2276 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:49 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2276 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:50 IST 2014 ************* total used free shared buffers cached Mem: 15759 4604 11155 0 24 2303 -/+ buffers/cache: 2276 13483 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:51 IST 2014 ************* total used free shared buffers cached Mem: 15759 4606 11153 0 24 2303 -/+ buffers/cache: 2277 13482 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:52 IST 2014 ************* total used free shared buffers cached Mem: 15759 4605 11153 0 24 2303 -/+ buffers/cache: 2277 13482 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:53 IST 2014 ************* total used free shared buffers cached Mem: 15759 4605 11153 0 24 2303 -/+ buffers/cache: 2277 13482 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:54 IST 2014 ************* total used free shared buffers cached Mem: 15759 4605 11153 0 24 2303 -/+ buffers/cache: 2277 13482 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:55 IST 2014 ************* total used free shared buffers cached Mem: 15759 4605 11154 0 24 2303 -/+ buffers/cache: 2277 13482 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:56 IST 2014 ************* total used free shared buffers cached Mem: 15759 4605 11154 0 24 2303 -/+ buffers/cache: 2277 13482 Swap: 1023 0 1023 *********** Tue Nov 25 18:06:57 IST 2014 ************* ^C [root@server2 ~]#
Regards,
Nagendra Rai
Created 11-25-2014 09:29 AM
Hi,
The explain is not really usefull for identifying the memory offenders. Can you post the output of profile (http://www.cloudera.com/content/cloudera/en/documentation/cloudera-impala/latest/topics/impala_expla... after you've run the query? That will shed some light on the actual memory requirements of each execution node in your query.
Thanks
Dimitris
Created 11-25-2014 09:50 AM
Hi,
In earlier post I have provided query profile details, I was not able to post whole output because it was crossing 50000 word msg limit and here we dont have any file attached option.
If you want I will provide the details again.
One more thing I would like to tell, when I am running this query my server RAM utilization never goes beyond 4.6 GB, then why impala is asking for 14GB memory and crashing if that much memory is not available.
Regards,
Nagendra Rai