Support Questions
Find answers, ask questions, and share your expertise
Announcements
Alert: Welcome to the Unified Cloudera Community. Former HCC members be sure to read and learn how to activate your account here.

NIFI putHiveStreaming processor error

Solved Go to solution

NIFI putHiveStreaming processor error

Master Collaborator

db-hive-temp.xml

nifi-db-hive.jpg

please see the attached error , i am feeding it JSON data but its complaining about AVRO format.

1 ACCEPTED SOLUTION

Accepted Solutions
Highlighted

Re: NIFI putHiveStreaming processor error

Super Guru
@Sami Ahmad

Output of QueryDatabaseTable processor are always in avro format , so you need to use PutHiveStreaming processor after Querydatabasetable processor.

As PutHiveStreaming processor expects incoming data to be in avro format and we are getting incoming data from querydatabasetable in avro format.

Flow:-

1.QueryDatabasetable
2.PutHiveStreaming
3.LogAttribute

please refer to below links regarding table creation of puthivestreaming processor

https://community.hortonworks.com/questions/59411/how-to-use-puthivestreaming.html
https://community.hortonworks.com/articles/52856/stream-data-into-hive-like-a-king-using-nifi.html

View solution in original post

15 REPLIES 15
Highlighted

Re: NIFI putHiveStreaming processor error

Hi @Sami Ahmad,

As stated in the processor description/documentation:

This processor uses Hive Streaming to send flow file data to an Apache Hive table. The incoming flow file is expected to be in Avro format and the table must exist in Hive. Please see the Hive documentation for requirements on the Hive table (format, partitions, etc.). The partition values are extracted from the Avro record based on the names of the partition columns as specified in the processor. NOTE: If multiple concurrent tasks are configured for this processor, only one table can be written to at any time by a single thread. Additional tasks intending to write to the same table will wait for the current task to finish writing to the table.

You'll need to convert your data into avro first. The best approach is to use the record processors for that.

Hope this helps.

Highlighted

Re: NIFI putHiveStreaming processor error

Master Collaborator
Shu this flow you mentioned 1) QueryDatabasetable 2)PutHiveStreaming 3)LogAttribute is not working for me , can you please take a look at the errors / the attached flow file ?
Highlighted

Re: NIFI putHiveStreaming processor error

Super Guru
@Sami Ahmad

Output of QueryDatabaseTable processor are always in avro format , so you need to use PutHiveStreaming processor after Querydatabasetable processor.

As PutHiveStreaming processor expects incoming data to be in avro format and we are getting incoming data from querydatabasetable in avro format.

Flow:-

1.QueryDatabasetable
2.PutHiveStreaming
3.LogAttribute

please refer to below links regarding table creation of puthivestreaming processor

https://community.hortonworks.com/questions/59411/how-to-use-puthivestreaming.html
https://community.hortonworks.com/articles/52856/stream-data-into-hive-like-a-king-using-nifi.html

View solution in original post

Highlighted

Re: NIFI putHiveStreaming processor error

Master Collaborator

Shu this flow you mentioned 1) QueryDatabasetable 2)PutHiveStreaming 3)LogAttribute is not working for me , can you please take a look at the errors / the attached flow file ?

Highlighted

Re: NIFI putHiveStreaming processor error

Super Guru
@Sami Ahmad

Sad to hear that

I tried to create the same flow and it works as expected.

Here is what i tried..

i have create a table in hive bucketed,partitioned,orc format,transactional.

hive ddl:-

hive> create table default.tmp(id string,name string)
partitioned by(ts timestamp)
clustered by (id)
into 2 buckets
stored as orc
TBLPROPERTIES ("transactional"="true");
hive> desc tmp;
+--------------------------+-----------------------+-----------------------+--+
|         col_name         |       data_type       |        comment        |
+--------------------------+-----------------------+-----------------------+--+
| id                       | string                |                       |
| name                     | string                |                       |
| ts                       | timestamp             |                       |
|                          | NULL                  | NULL                  |
| # Partition Information  | NULL                  | NULL                  |
| # col_name               | data_type             | comment               |
|                          | NULL                  | NULL                  |
| ts                       | timestamp             |                       |
+--------------------------+-----------------------+-----------------------+--+

i have created the tmp table with id,name columns as string types and partitioned by ts as timestamp type.

Now PutHiveStreaming Processor Configs:-

62862-puthivestreaming.png

select data from tmp table now:-

hive> select * from tmp;
+-----+-------+--------------------------+--+
| id  | name  |            ts            |
+-----+-------+--------------------------+--+
| 1   | hdf   | 2015-01-02 20:23:20.0    |
| 1   | hdf   | 2015-01-02 20:23:20.0    |
| 1   | hdf   | 2018-03-12 19:11:51.217  |
| 1   | hdf   | 2018-03-12 19:11:51.24   |
+-----+-------+--------------------------+--+
hive> show partitions tmp;
+---------------------------------+--+
|            partition            |
+---------------------------------+--+
| ts=2015-01-02 20%3A23%3A20.0    |
| ts=2018-03-12 19%3A11%3A51.217  |
| ts=2018-03-12 19%3A11%3A51.24   |
+---------------------------------+--+

For testing purpose use select hiveql processor with below configs:-

HiveQL Select Query

select int(1) id,string('hdf')name,timestamp(current_timestamp)ts

Output Format

Avro

Configs:-

62863-selecthiveql.png

Test Flow:-

62864-hivestreaming-test-flow.png

try this flow once and make sure you are able to load using PutHiveStreaming processor into hive.once you are able to load the data into hive then try with your scenario.

If issue still exists then share your hive create table statement and 10 records of sample data so that we can recreate and help you out ASAP.

In addition if you don't want to create partitions then create table statement would be

hive> create table default.tmp(id string,name string,ts string) --with ts as timestamp type also works
clustered by (id)
into 2 buckets
stored as orc
TBLPROPERTIES ("transactional"="true");

62865-nopartition.png

Highlighted

Re: NIFI putHiveStreaming processor error

Master Collaborator

hi Shu can you please put the properties of the HiveConnectionPool ?

hiveconnectionpool-properties.jpg

Highlighted

Re: NIFI putHiveStreaming processor error

Super Guru
@Sami Ahmad

This hive connection pool configs are varies depending on the setup,unfortunately there are no standard properties that i can share with you.

HiveConnectionPool configs:-

I use the below connection pool configs
Database Connection URL

jdbc:hive2://sandbox.hortonworks:2181,sandbox.hortonworks:2181,sandbox.hortonworks:2181/;serviceDiscoveryMode=zooKeeper;zooKeeperNamespace=hiveserver2-binary;principal=hive/_HOST@HDP.LOCAL

Hive Configuration Resources

/etc/hdp/hive-site.xml,/etc/hdp/core-site.xml,/etc/hdp/hdfs-site.xml

Database User

nifi

Password No value set Max Wait Time

500 millis

Max Total Connections

8

Validation query

select current_timestamp

Please refer to below links to get your jdbc connection url and if your environment is kerberized then you need to mention kerberos principal,keytab in the connection pool service.

https://community.hortonworks.com/questions/107945/hive-database-connection-pooling-service.html

https://community.hortonworks.com/articles/4103/hiveserver2-jdbc-connection-url-examples.html
https://community.hortonworks.com/questions/64609/cannot-create-hive-connection-pool.html

Highlighted

Re: NIFI putHiveStreaming processor error

Super Guru
@Sami Ahmad

You have to change the data types in your create table statement as table used by put hive streaming processor expects all the datatypes would be strings(except for timestamp..etc types).

use the below create table statement

CREATE TABLE default.purchase_acct_orc (
acct_num STRING,
pur_id STRING,
pur_det_id STRING,
product_pur_product_code STRING,
prod_amt STRING,
accttype_acct_type_code STRING,
accttype_acct_status_code STRING,
emp_emp_code STRING,
plaza_plaza_id STRING,
purstat_pur_status_code STRING)
PARTITIONED BY (pur_trans_date TIMESTAMP)
CLUSTERED BY(acct_num) INTO 5 BUCKETS
STORED AS ORC
TBLPROPERTIES ("transactional"="true");

Then start the put hive streaming processor.

As i tried with same create table statement as above and the data that you have provided in the comment and i'm able to load the data into the table.

Flow:-

64734-flow.png

Select Hive Ql configs:-

64733-selecthiveql.png

HiveQL Select Query

select string(16680934) ACCT_NUM,string(747032990) PUR_ID,string(656564856) PUR_DET_ID,string(31) PRODUCT_PUR_PRODUCT_CODE,string(50) PROD_AMT,string('2015-01-01 08:12:03.0') PUR_TRANS_DATE,string(1) ACCTTYPE_ACCT_TYPE_CODE,string(01) ACCTTYPE_ACCT_STATUS_CODE,string(9999) EMP_EMP_CODE,string(009500 )PLAZA_PLAZA_ID,string(10) PURSTAT_PUR_STATUS_CODE

Output Format

Avro

Puthivestreaming Configs:-

64732-hivestreaming.png

Output:-

 select * from default.purchase_acct_orc;
+-----------+------------+-------------+---------------------------+-----------+--------------------------+----------------------------+---------------+-----------------+--------------------------+------------------------+--+
| acct_num  |   pur_id   | pur_det_id  | product_pur_product_code  | prod_amt  | accttype_acct_type_code  | accttype_acct_status_code  | emp_emp_code  | plaza_plaza_id  | purstat_pur_status_code  |     pur_trans_date     |
+-----------+------------+-------------+---------------------------+-----------+--------------------------+----------------------------+---------------+-----------------+--------------------------+------------------------+--+
| 16680934  | 747032990  | 656564856   | 31                        | 50        | 1                        | 1                          | 9999          | 9500            | 10                       | 2015-01-01 08:12:03.0  |
+-----------+------------+-------------+---------------------------+-----------+--------------------------+----------------------------+---------------+-----------------+--------------------------+------------------------+--+

Let us know if you have any issues..!!

Highlighted

Re: NIFI putHiveStreaming processor error

Master Collaborator

hi Shu

yes the table structure you gave fixed the issue and its loading now but something strange is happening , it got stuck on 104 for a while and I stopped the QueryDatabaseTable processor , then after a while I noticed it had dumped more records . I guess its just slow ?

how can I tune the loading app as its extremely slow.

thanks

18555604        747495984       657020052       31      25      01      NULL    9985    009500  10      2015-01-03 10:02:41
Time taken: 0.089 seconds, Fetched: 104 row(s)
hive>
<br>
Don't have an account?
Coming from Hortonworks? Activate your account here