Question

我正试图在pyspark的apache phoenix中按日期过滤。凤凰中的列创建为Date，过滤器是日期时间。当我使用解释时，我看到火花不会将过滤器推向凤凰。我没有运气就尝试了很多组合。

有什么办法吗？

df = sqlContext.read \
   .format("org.apache.phoenix.spark") \
  .option("table", "TABLENAME") \
  .option("zkUrl",zookepperServer +":2181:/hbase-unsecure" ) \
  .load()
print(df.printSchema())

startValidation = datetime.datetime.now()

print(df.filter(df['FH'] >startValidation).explain(True))

结果

root
 |-- METER_ID: string (nullable = true)
 |-- FH: date (nullable = true)

None
   == Parsed Logical Plan ==
'Filter (FH#53 > 1486726683446150)
+- Relation[METER_ID#52,FH#53,SUMMERTIME#54,MAGNITUDE#55,SOURCE#56,ENTRY_DATETIME#57,BC#58,T_VAL_AE#59,T_VAL_AI#60,T_VAL_R1#61,T_VAL_R2#62,T_VAL_R3#63,T_VAL_R4#64] PhoenixRelation(DAILYREADS,10.0.0.13:2181:/hbase-unsecure)

== Analyzed Logical Plan ==
METER_ID: string, FH: date, SUMMERTIME: string, MAGNITUDE: int, SOURCE: int, ENTRY_DATETIME: date, BC: string, T_VAL_AE: int, T_VAL_AI: int, T_VAL_R1: int, T_VAL_R2: int, T_VAL_R3: int, T_VAL_R4: int
Filter (cast(FH#53 as string) > cast(1486726683446150 as string))
+- Relation[METER_ID#52,FH#53,SUMMERTIME#54,MAGNITUDE#55,SOURCE#56,ENTRY_DATETIME#57,BC#58,T_VAL_AE#59,T_VAL_AI#60,T_VAL_R1#61,T_VAL_R2#62,T_VAL_R3#63,T_VAL_R4#64] PhoenixRelation(DAILYREADS,10.0.0.13:2181:/hbase-unsecure)

== Optimized Logical Plan ==
Filter (cast(FH#53 as string) > 2017-02-10 11:38:03.44615)
+- Relation[METER_ID#52,FH#53,SUMMERTIME#54,MAGNITUDE#55,SOURCE#56,ENTRY_DATETIME#57,BC#58,T_VAL_AE#59,T_VAL_AI#60,T_VAL_R1#61,T_VAL_R2#62,T_VAL_R3#63,T_VAL_R4#64] PhoenixRelation(DAILYREADS,10.0.0.13:2181:/hbase-unsecure)

== Physical Plan ==
Filter (cast(FH#53 as string) > 2017-02-10 11:38:03.44615)
+- Scan PhoenixRelation(DAILYREADS,10.0.0.13:2181:/hbase-unsecure)[METER_ID#52,FH#53,SUMMERTIME#54,MAGNITUDE#55,SOURCE#56,ENTRY_DATETIME#57,BC#58,T_VAL_AE#59,T_VAL_AI#60,T_VAL_R1#61,T_VAL_R2#62,T_VAL_R3#63,T_VAL_R4#64]
None

如果我将FH列设置为时间戳，则会推送过滤器但会引发异常：

Caused by: org.apache.phoenix.exception.PhoenixParserException: ERROR 604 (42P00): Syntax error. Mismatched input. Expecting "RPAREN", got "12" at line 1, column 219.
    at org.apache.phoenix.exception.PhoenixParserException.newException(PhoenixParserException.java:33)
    at org.apache.phoenix.parse.SQLParser.parseStatement(SQLParser.java:111)
    at org.apache.phoenix.jdbc.PhoenixStatement$PhoenixStatementParser.parseStatement(PhoenixStatement.java:1280)
    at org.apache.phoenix.jdbc.PhoenixStatement.parseStatement(PhoenixStatement.java:1363)
    at org.apache.phoenix.jdbc.PhoenixStatement.compileQuery(PhoenixStatement.java:1373)
    at org.apache.phoenix.jdbc.PhoenixStatement.optimizeQuery(PhoenixStatement.java:1368)
    at org.apache.phoenix.mapreduce.PhoenixInputFormat.getQueryPlan(PhoenixInputFormat.java:122)
    ... 102 more
Caused by: MismatchedTokenException(106!=129)
    at org.apache.phoenix.parse.PhoenixSQLParser.recoverFromMismatchedToken(PhoenixSQLParser.java:360)
    at org.apache.phoenix.shaded.org.antlr.runtime.BaseRecognizer.match(BaseRecognizer.java:115)
    at org.apache.phoenix.parse.PhoenixSQLParser.not_expression(PhoenixSQLParser.java:6862)
    at org.apache.phoenix.parse.PhoenixSQLParser.and_expression(PhoenixSQLParser.java:6677)
    at org.apache.phoenix.parse.PhoenixSQLParser.or_expression(PhoenixSQLParser.java:6614)
    at org.apache.phoenix.parse.PhoenixSQLParser.expression(PhoenixSQLParser.java:6579)
    at org.apache.phoenix.parse.PhoenixSQLParser.single_select(PhoenixSQLParser.java:4615)
    at org.apache.phoenix.parse.PhoenixSQLParser.unioned_selects(PhoenixSQLParser.java:4697)
    at org.apache.phoenix.parse.PhoenixSQLParser.select_node(PhoenixSQLParser.java:4763)
    at org.apache.phoenix.parse.PhoenixSQLParser.oneStatement(PhoenixSQLParser.java:789)
    at org.apache.phoenix.parse.PhoenixSQLParser.statement(PhoenixSQLParser.java:508)
    at org.apache.phoenix.parse.SQLParser.parseStatement(SQLParser.java:108)
    ... 107 more

由于

Answer 1

通常会在Apache开发人员列表中询问问题，并在JIRA中报告问题。您的JIRA在4小时内得到了回复：https://issues.apache.org/jira/browse/PHOENIX-3664?focusedCommentId=15863720&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-15863720

不是霰弹枪+堆栈溢出+开发列表+ JIRA的霰弹枪方法的忠实粉丝。请记住，所有回复的人都有日常工作。

Pyspark：按照日期对阿帕奇凤凰推进过滤器

1 个答案: