[DRILL-7953] Query failed with (Too many open files)

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Not A Bug
• Affects Version/s: 1.15.0
• Component/s: Server
• Assignee: Unassigned
• Reporter: Dony Dong
• Created: 21/Jun/21 05:27
• Updated: 29/May/22 07:27
• Resolved: 29/May/22 07:27

Description

Hi Support,

 

When we query a complex view that will access a lot of parquet files, the query failed with error,  

Caused by: org.apache.drill.common.exceptions.ExecutionSetupException: Error opening or reading metadata for parquet file at location: part-00006-df5fe7db-6086-43a3-9575-1b18c140b5e6-c000.snappy.parquet
at org.apache.drill.exec.store.parquet.columnreaders.PageReader.<init>(PageReader.java:151) ~[drill-java-exec-1.15.0.jar:1.15.0]
at org.apache.drill.exec.store.parquet.columnreaders.AsyncPageReader.<init>(AsyncPageReader.java:97) ~[drill-java-exec-1.15.0.jar:1.15.0]
at org.apache.drill.exec.store.parquet.columnreaders.ColumnReader.<init>(ColumnReader.java:100) ~[drill-java-exec-1.15.0.jar:1.15.0]
at org.apache.drill.exec.store.parquet.columnreaders.NullableColumnReader.<init>(NullableColumnReader.java:43) ~[drill-java-exec-1.15.0.jar:1.15.0]
at org.apache.drill.exec.store.parquet.columnreaders.NullableFixedByteAlignedReaders$NullableFixedByteAlignedReader.<init>(NullableFixedByteAlignedReaders.java:54) ~[drill-java-exec-1.15.0.jar:1.15.0]
at org.apache.drill.exec.store.parquet.columnreaders.NullableFixedByteAlignedReaders$NullableConvertedReader.<init>(NullableFixedByteAlignedReaders.java:328) ~[drill-java-exec-1.15.0.jar:1.15.0]
at org.apache.drill.exec.store.parquet.columnreaders.NullableFixedByteAlignedReaders$NullableDateReader.<init>(NullableFixedByteAlignedReaders.java:348) ~[drill-java-exec-1.15.0.jar:1.15.0]
at org.apache.drill.exec.store.parquet.columnreaders.ColumnReaderFactory.createFixedColumnReader(ColumnReaderFactory.java:185) ~[drill-java-exec-1.15.0.jar:1.15.0]
at org.apache.drill.exec.store.parquet.columnreaders.ParquetColumnMetadata.makeFixedWidthReader(ParquetColumnMetadata.java:141) ~[drill-java-exec-1.15.0.jar:1.15.0]
at org.apache.drill.exec.store.parquet.columnreaders.ReadState.buildReader(ReadState.java:123) ~[drill-java-exec-1.15.0.jar:1.15.0]
at org.apache.drill.exec.store.parquet.columnreaders.ParquetRecordReader.setup(ParquetRecordReader.java:253) ~[drill-java-exec-1.15.0.jar:1.15.0]
... 29 common frames omitted
Caused by: java.io.FileNotFoundException: /data/testing/CH/part-00006-df5fe7db-6086-43a3-9575-1b18c140b5e6-c000.snappy.parquet (Too many open files)
at java.io.FileInputStream.open0(Native Method) ~[na:1.8.0_181]
at java.io.FileInputStream.open(FileInputStream.java:195) ~[na:1.8.0_181]
at java.io.FileInputStream.<init>(FileInputStream.java:138) ~[na:1.8.0_181]
at org.apache.hadoop.fs.RawLocalFileSystem$LocalFSFileInputStream.<init>(RawLocalFileSystem.java:106) ~[hadoop-common-2.7.4.jar:na]
at org.apache.hadoop.fs.RawLocalFileSystem.open(RawLocalFileSystem.java:202) ~[hadoop-common-2.7.4.jar:na]
at org.apache.hadoop.fs.ChecksumFileSystem$ChecksumFSInputChecker.<init>(ChecksumFileSystem.java:143) ~[hadoop-common-2.7.4.jar:na]
at org.apache.hadoop.fs.ChecksumFileSystem.open(ChecksumFileSystem.java:346) ~[hadoop-common-2.7.4.jar:na]
at org.apache.hadoop.fs.FileSystem.open(FileSystem.java:769) ~[hadoop-common-2.7.4.jar:na]
at org.apache.drill.exec.store.dfs.DrillFileSystem.open(DrillFileSystem.java:151) ~[drill-java-exec-1.15.0.jar:1.15.0]
at org.apache.drill.exec.store.parquet.columnreaders.PageReader.<init>(PageReader.java:133) ~[drill-java-exec-1.15.0.jar:1.15.0]
... 39 common frames omitted

 

We add below in the /etc/security/limits.conf, but Drill still uses the default setting 1024 when startup. 

• hard nofile 65536
• soft nofile 65536

 

Fri Jun 18 19:40:33 AEST 2021 Starting drillbit on drill-testing
core file size (blocks, -c) 0
data seg size (kbytes, -d) unlimited
scheduling priority (-e) 0
file size (blocks, -f) unlimited
pending signals (-i) 740731
max locked memory (kbytes, -l) 64
max memory size (kbytes, -m) unlimited
open files (-n) 1024
pipe size (512 bytes, -p) 8
POSIX message queues (bytes, -q) 819200
real-time priority (-r) 0
stack size (kbytes, -s) 8192
cpu time (seconds, -t) unlimited
max user processes (-u) 740731
virtual memory (kbytes, -v) unlimited
file locks (-x) unlimited

 

is there some place we can set this parameter?