Utilizing Apache Pulsar to Populate Apache Iceberg and Apache Parquet based Lakehouses
FLiP-Pi-Iceberg-Thermal
Apache Iceberg + Apache Pulsar + Thermal Sensor Data from a Raspberry Pi
Steps
- Run Apache Pulsar 2.10.2 (standalone, docker, baremetal cluster, VM cluster, K8 cluster, AWS Marketplace Pulsar, StreamNative Cloud)
- Run Apache Iceberg (docker, ...) 1.1.0
- Run Apache Spark 3.2
- Deploy Pulsar connector
- Send data to Pulsar topic
- Query Iceberg in Spark
Sensor Python App Sending messages
{'uuid': 'thrml_wse_20221216202136', 'ipaddress': '192.168.1.179', 'cputempf': 122, 'runtime': 0, 'host': 'thermal', 'hostname': 'thermal', 'macaddress': 'e4:5f:01:7c:3f:34', 'endtime': '1671222096.350368', 'te': '0.0005612373352050781', 'cpu': 5.5, 'diskusage': '101858.0 MB', 'memory': 9.9, 'rowid': '20221216202136_a0e9eae8-3b4f-4222-95c6-7657ba0e12e2', 'systemtime': '12/16/2022 15:21:41', 'ts': 1671222101, 'starttime': '12/16/2022 15:21:36', 'datetimestamp': '2022-12-16 20:21:40.012859+00:00', 'temperature': 30.5959, 'humidity': 26.07, 'co2': 767.0, 'totalvocppb': 0.0, 'equivalentco2ppm': 400.0, 'pressure': 99773.53, 'temperatureicp': 86.0}
Pulsar Sink Deploy
bin/pulsar-admin sink stop --name iceberg_sink --namespace default --tenant public
bin/pulsar-admin sinks delete --tenant public --namespace default --name iceberg_sink
bin/pulsar-admin sink create --sink-config-file conf/iceberg.json
Pulsar Sink Status
bin/pulsar-admin sinks status --tenant public --namespace default --name iceberg_sink
{
"numInstances" : 1,
"numRunning" : 1,
"instances" : [ {
"instanceId" : 0,
"status" : {
"running" : true,
"error" : "",
"numRestarts" : 0,
"numReadFromPulsar" : 10,
"numSystemExceptions" : 0,
"latestSystemExceptions" : [ ],
"numSinkExceptions" : 0,
"latestSinkExceptions" : [ ],
"numWrittenToSink" : 10,
"lastReceivedTime" : 1671220772536,
"workerId" : "c-standalone-fw-127.0.0.1-8080"
}
} ]
}
Iceberg data written via Pulsar Lakehouse Cloud Sink
ls -lt /Users/tspann/Downloads/iceberg/iceberg_sink_test/ice_sink_thermal
total 0
drwxr-xr-x 94 tspann staff 3008 Dec 16 15:45 metadata
drwxr-xr-x 34 tspann staff 1088 Dec 16 15:45 data
ice_sink_thermal/metadata
total 856
-rw-r--r-- 1 tspann staff 2 Dec 16 15:45 version-hint.text
-rw-r--r-- 1 tspann staff 19283 Dec 16 15:45 v15.metadata.json
-rw-r--r-- 1 tspann staff 4352 Dec 16 15:45 snap-8802315029762513718-1-78627844-0d69-4c2e-87db-016b9fdac119.avro
-rw-r--r-- 1 tspann staff 7536 Dec 16 15:45 78627844-0d69-4c2e-87db-016b9fdac119-m0.avro
-rw-r--r-- 1 tspann staff 18303 Dec 16 15:43 v14.metadata.json
-rw-r--r-- 1 tspann staff 4315 Dec 16 15:43 snap-1218246990201737819-1-1253a40d-fae5-4919-9d71-be51af402899.avro
iceberg_sink_test/ice_sink_thermal/data
total 360
-rw-r--r-- 1 tspann staff 9771 Dec 16 15:45 00000-1-951a37fc-5069-4201-94fa-4ef9975f6293-00014.parquet
-rw-r--r-- 1 tspann staff 9782 Dec 16 15:43 00000-1-951a37fc-5069-4201-94fa-4ef9975f6293-00013.parquet
-rw-r--r-- 1 tspann staff 9733 Dec 16 15:41 00000-1-951a37fc-5069-4201-94fa-4ef9975f6293-00012.parquet
-rw-r--r-- 1 tspann staff 9637 Dec 16 15:39 00000-1-951a37fc-5069-4201-94fa-4ef9975f6293-00011.parquet
-rw-r--r-- 1 tspann staff 9722 Dec 16 15:37 00000-1-951a37fc-5069-4201-94fa-4ef9975f6293-00010.parquet
-rw-r--r-- 1 tspann staff 9663 Dec 16 15:35 00000-1-951a37fc-5069-4201-94fa-4ef9975f6293-00009.parquet
-rw-r--r-- 1 tspann staff 9671 Dec 16 15:33 00000-1-951a37fc-5069-4201-94fa-4ef9975f6293-00008.parquet
-rw-r--r-- 1 tspann staff 9652 Dec 16 15:31 00000-1-951a37fc-5069-4201-94fa-4ef9975f6293-00007.parquet
-rw-r--r-- 1 tspann staff 9716 Dec 16 15:29 00000-1-951a37fc-5069-4201-94fa-4ef9975f6293-00006.parquet
-rw-r--r-- 1 tspann staff 9731 Dec 16 15:27 00000-1-951a37fc-5069-4201-94fa-4ef9975f6293-00005.parquet
-rw-r--r-- 1 tspann staff 9639 Dec 16 15:25 00000-1-951a37fc-5069-4201-94fa-4ef9975f6293-00004.parquet
-rw-r--r-- 1 tspann staff 9721 Dec 16 15:23 00000-1-951a37fc-5069-4201-94fa-4ef9975f6293-00003.parquet
-rw-r--r-- 1 tspann staff 7414 Dec 16 15:21 00000-1-951a37fc-5069-4201-94fa-4ef9975f6293-00002.parquet
-rw-r--r-- 1 tspann staff 8492 Dec 16 14:59 00000-1-951a37fc-5069-4201-94fa-4ef9975f6293-00001.parquet
-rw-r--r-- 1 tspann staff 7978 Dec 16 14:56 00000-1-2acfc6ba-4f49-44c7-8f17-1f3491484fd1-00001.parquet
-rw-r--r-- 1 tspann staff 7886 Dec 16 14:54 00000-1-7e714ed7-0ba5-41a4-b8e1-1e1d261e3b83-00001.parquet
Schema Embedded in Parquet File
{"type":"struct","schema-id":0,"fields":[{"id":1,"name":"uuid","required":true,"type":"string"},{"id":2,"name":"ipaddress","required":true,"type":"string"},{"id":3,"name":"cputempf","required":true,"type":"int"},{"id":4,"name":"runtime","required":true,"type":"int"},{"id":5,"name":"host","required":true,"type":"string"},{"id":6,"name":"hostname","required":true,"type":"string"},{"id":7,"name":"macaddress","required":true,"type":"string"},{"id":8,"name":"endtime","required":true,"type":"string"},{"id":9,"name":"te","required":true,"type":"string"},{"id":10,"name":"cpu","required":true,"type":"float"},{"id":11,"name":"diskusage","required":true,"type":"string"},{"id":12,"name":"memory","required":true,"type":"float"},{"id":13,"name":"rowid","required":true,"type":"string"},{"id":14,"name":"systemtime","required":true,"type":"string"},{"id":15,"name":"ts","required":true,"type":"int"},{"id":16,"name":"starttime","required":true,"type":"string"},{"id":17,"name":"datetimestamp","required":true,"type":"string"},{"id":18,"name":"temperature","required":true,"type":"float"},{"id":19,"name":"humidity","required":true,"type":"float"},{"id":20,"name":"co2","required":true,"type":"float"},{"id":21,"name":"totalvocppb","required":true,"type":"float"},{"id":22,"name":"equivalentco2ppm","required":true,"type":"float"},{"id":23,"name":"pressure","required":true,"type":"float"},{"id":24,"name":"temperatureicp","required":true,"type":"float"}]}Jparquet-mr version 1.12.0 (build db75a6815f2ba1d1ee89d1a90aeb296f1f3a8f20)
Validate our Parquet Files
pip3.9 install parquet-tools -U
parquet-tools inspect ice_sink_thermal/data/00000-1-7e714ed7-0ba5-41a4-b8e1-1e1d261e3b83-00001.parquet
############ file meta data ############
created_by: parquet-mr version 1.12.0 (build db75a6815f2ba1d1ee89d1a90aeb296f1f3a8f20)
num_columns: 24
num_rows: 4
num_row_groups: 1
format_version: 1.0
serialized_size: 4577
############ Columns ############
uuid
ipaddress
cputempf
runtime
host
hostname
macaddress
endtime
te
cpu
diskusage
memory
rowid
systemtime
ts
starttime
datetimestamp
temperature
humidity
co2
totalvocppb
equivalentco2ppm
pressure
temperatureicp
############ Column(uuid) ############
name: uuid
path: uuid
max_definition_level: 0
max_repetition_level: 0
physical_type: BYTE_ARRAY
logical_type: String
converted_type (legacy): UTF8
compression: GZIP (space_saved: 28%)
############ Column(ipaddress) ############
name: ipaddress
path: ipaddress
max_definition_level: 0
max_repetition_level: 0
physical_type: BYTE_ARRAY
logical_type: String
converted_type (legacy): UTF8
compression: GZIP (space_saved: -67%)
############ Column(cputempf) ############
name: cputempf
path: cputempf
max_definition_level: 0
max_repetition_level: 0
physical_type: INT32
logical_type: None
converted_type (legacy): NONE
compression: GZIP (space_saved: -38%)
############ Column(runtime) ############
name: runtime
path: runtime
max_definition_level: 0
max_repetition_level: 0
physical_type: INT32
logical_type: None
converted_type (legacy): NONE
compression: GZIP (space_saved: -85%)
############ Column(host) ############
name: host
path: host
max_definition_level: 0
max_repetition_level: 0
physical_type: BYTE_ARRAY
logical_type: String
converted_type (legacy): UTF8
compression: GZIP (space_saved: -74%)
############ Column(hostname) ############
name: hostname
path: hostname
max_definition_level: 0
max_repetition_level: 0
physical_type: BYTE_ARRAY
logical_type: String
converted_type (legacy): UTF8
compression: GZIP (space_saved: -74%)
############ Column(macaddress) ############
name: macaddress
path: macaddress
max_definition_level: 0
max_repetition_level: 0
physical_type: BYTE_ARRAY
logical_type: String
converted_type (legacy): UTF8
compression: GZIP (space_saved: -62%)
############ Column(endtime) ############
name: endtime
path: endtime
max_definition_level: 0
max_repetition_level: 0
physical_type: BYTE_ARRAY
logical_type: String
converted_type (legacy): UTF8
compression: GZIP (space_saved: 17%)
############ Column(te) ############
name: te
path: te
max_definition_level: 0
max_repetition_level: 0
physical_type: BYTE_ARRAY
logical_type: String
converted_type (legacy): UTF8
compression: GZIP (space_saved: 16%)
############ Column(cpu) ############
name: cpu
path: cpu
max_definition_level: 0
max_repetition_level: 0
physical_type: FLOAT
logical_type: None
converted_type (legacy): NONE
compression: GZIP (space_saved: -75%)
############ Column(diskusage) ############
name: diskusage
path: diskusage
max_definition_level: 0
max_repetition_level: 0
physical_type: BYTE_ARRAY
logical_type: String
converted_type (legacy): UTF8
compression: GZIP (space_saved: -69%)
############ Column(memory) ############
name: memory
path: memory
max_definition_level: 0
max_repetition_level: 0
physical_type: FLOAT
logical_type: None
converted_type (legacy): NONE
compression: GZIP (space_saved: -85%)
############ Column(rowid) ############
name: rowid
path: rowid
max_definition_level: 0
max_repetition_level: 0
physical_type: BYTE_ARRAY
logical_type: String
converted_type (legacy): UTF8
compression: GZIP (space_saved: 33%)
############ Column(systemtime) ############
name: systemtime
path: systemtime
max_definition_level: 0
max_repetition_level: 0
physical_type: BYTE_ARRAY
logical_type: String
converted_type (legacy): UTF8
compression: GZIP (space_saved: 33%)
############ Column(ts) ############
name: ts
path: ts
max_definition_level: 0
max_repetition_level: 0
physical_type: INT32
logical_type: None
converted_type (legacy): NONE
compression: GZIP (space_saved: -41%)
############ Column(starttime) ############
name: starttime
path: starttime
max_definition_level: 0
max_repetition_level: 0
physical_type: BYTE_ARRAY
logical_type: String
converted_type (legacy): UTF8
compression: GZIP (space_saved: 33%)
############ Column(datetimestamp) ############
name: datetimestamp
path: datetimestamp
max_definition_level: 0
max_repetition_level: 0
physical_type: BYTE_ARRAY
logical_type: String
converted_type (legacy): UTF8
compression: GZIP (space_saved: 37%)
############ Column(temperature) ############
name: temperature
path: temperature
max_definition_level: 0
max_repetition_level: 0
physical_type: FLOAT
logical_type: None
converted_type (legacy): NONE
compression: GZIP (space_saved: -51%)
############ Column(humidity) ############
name: humidity
path: humidity
max_definition_level: 0
max_repetition_level: 0
physical_type: FLOAT
logical_type: None
converted_type (legacy): NONE
compression: GZIP (space_saved: -49%)
############ Column(co2) ############
name: co2
path: co2
max_definition_level: 0
max_repetition_level: 0
physical_type: FLOAT
logical_type: None
converted_type (legacy): NONE
compression: GZIP (space_saved: -51%)
############ Column(totalvocppb) ############
name: totalvocppb
path: totalvocppb
max_definition_level: 0
max_repetition_level: 0
physical_type: FLOAT
logical_type: None
converted_type (legacy): NONE
compression: GZIP (space_saved: -85%)
############ Column(equivalentco2ppm) ############
name: equivalentco2ppm
path: equivalentco2ppm
max_definition_level: 0
max_repetition_level: 0
physical_type: FLOAT
logical_type: None
converted_type (legacy): NONE
compression: GZIP (space_saved: -75%)
############ Column(pressure) ############
name: pressure
path: pressure
max_definition_level: 0
max_repetition_level: 0
physical_type: FLOAT
logical_type: None
converted_type (legacy): NONE
compression: GZIP (space_saved: -49%)
############ Column(temperatureicp) ############
name: temperatureicp
path: temperatureicp
max_definition_level: 0
max_repetition_level: 0
physical_type: FLOAT
logical_type: None
converted_type (legacy): NONE
compression: GZIP (space_saved: -85%)
Setup
- Download Spark 3.2_2.12
- Download iceberg-spark-runtime-3.2_2.12:1.1.0
Run Spark Shell
bin/spark-sql --packages org.apache.iceberg:iceberg-spark-runtime-3.2_2.12:1.1.0\
--conf spark.sql.extensions=org.apache.iceberg.spark.extensions.IcebergSparkSessionExtensions \
--conf spark.sql.catalog.spark_catalog=org.apache.iceberg.spark.SparkSessionCatalog \
--conf spark.sql.catalog.spark_catalog.type=hive \
--conf spark.sql.catalog.local=org.apache.iceberg.spark.SparkCatalog \
--conf spark.sql.catalog.local.type=hadoop \
--conf spark.sql.catalog.local.warehouse=/Users/tspann/Downloads/iceberg/iceberg_sink_test
Spark Shell
desc local.ice_sink_thermal;
uuid string
ipaddress string
cputempf int
runtime int
host string
hostname string
macaddress string
endtime string
te string
cpu float
diskusage string
memory float
rowid string
systemtime string
ts int
starttime string
datetimestamp string
temperature float
humidity float
co2 float
totalvocppb float
equivalentco2ppm float
pressure float
temperatureicp float
# Partitioning
Not partitioned
select * from local.ice_sink_thermal limit 10;
thrml_zlq_20221216202731 192.168.1.179 122 0 thermal thermal e4:5f:01:7c:3f:34 1671222451.063736 0.0005979537963867188 11.1 101858.0 MB 10.0 20221216202731_e414e311-7928-4408-91de-b44666cd14db 12/16/2022 15:27:35 1671222455 12/16/2022 15:27:31 2022-12-16 20:27:34.827283+00:00 26.8868 31.76 771.0 14.0 405.0 99783.77 86.0
thrml_fpa_20221216202735 192.168.1.179 122 0 thermal thermal e4:5f:01:7c:3f:34 1671222455.8950574 0.00046181678771972656 5.5 101858.0 MB 10.0 20221216202735_92a2af14-ebcb-42fd-a935-5a01a47ff95e 12/16/2022 15:27:40 1671222460 12/16/2022 15:27:35 2022-12-16 20:27:39.659337+00:00 26.8761 31.74 771.0 6.0 400.0 99784.8 86.0
thrml_gpv_20221216202740 192.168.1.179 122 0 thermal thermal e4:5f:01:7c:3f:34 1671222460.7061012 0.0006053447723388672 5.5 101858.0 MB 10.0 20221216202740_68f88218-1b40-4f0a-85e3-3fb8f447d65b 12/16/2022 15:27:45 1671222465 12/16/2022 15:27:40 2022-12-16 20:27:44.369295+00:00 26.8761 31.72 770.0 7.0 65535.0 99782.18 85.0
thrml_gxu_20221216202745 192.168.1.179 121 0 thermal thermal e4:5f:01:7c:3f:34 1671222465.500708 0.0006241798400878906 6.0 101858.0 MB 10.0 20221216202745_28685c57-88f5-422b-be23-b32ea12a0d75 12/16/2022 15:27:50 1671222470 12/16/2022 15:27:45 2022-12-16 20:27:49.161722+00:00 26.8681 31.83 771.0 12.0 65535.0 99778.97 85.0
thrml_nyn_20221216202750 192.168.1.179 122 0 thermal thermal e4:5f:01:7c:3f:34 1671222470.212329 0.0006175041198730469 5.5 101858.0 MB 10.0 20221216202750_bbbb7fa7-cebf-4414-b538-30ce84828cea 12/16/2022 15:27:55 1671222475 12/16/2022 15:27:50 2022-12-16 20:27:53.976872+00:00 26.8601 31.78 771.0 6.0 65535.0 99783.55 86.0
thrml_oxl_20221216202755 192.168.1.179 123 0 thermal thermal e4:5f:01:7c:3f:34 1671222475.1313043 0.0006723403930664062 6.6 101858.0 MB 10.0 20221216202755_bda36e1c-246f-4d99-8b39-b558111a1d9e 12/16/2022 15:27:59 1671222479 12/16/2022 15:27:55 2022-12-16 20:27:58.794626+00:00 26.8307 31.73 771.0 7.0 65535.0 99785.26 86.0
thrml_rvg_20221216202759 192.168.1.179 121 0 thermal thermal e4:5f:01:7c:3f:34 1671222479.8391178 0.00048804283142089844 5.5 101858.0 MB 10.4 20221216202759_25b8ee3b-6b59-42e6-8c4c-a8419b76ea40 12/16/2022 15:28:04 1671222484 12/16/2022 15:27:59 2022-12-16 20:28:03.601685+00:00 26.8441 31.76 771.0 6.0 406.0 99786.36 86.0
thrml_wbl_20221216202804 192.168.1.179 123 0 thermal thermal e4:5f:01:7c:3f:34 1671222484.6672618 0.0006029605865478516 5.5 101858.0 MB 10.0 20221216202804_7bd3de16-dbcd-4107-8ab0-b184a2eaf523 12/16/2022 15:28:09 1671222489 12/16/2022 15:28:04 2022-12-16 20:28:08.431971+00:00 26.8842 31.79 770.0 9.0 65535.0 99787.27 86.0
thrml_vwj_20221216202809 192.168.1.179 122 0 thermal thermal e4:5f:01:7c:3f:34 1671222489.4752936 0.0006010532379150391 9.9 101858.0 MB 10.0 20221216202809_fed14b6d-b211-48ad-b31f-0a486b8d0f0d 12/16/2022 15:28:14 1671222494 12/16/2022 15:28:09 2022-12-16 20:28:13.137929+00:00 26.9002 31.78 770.0 5.0 400.0 99782.4 86.0
thrml_oox_20221216202814 192.168.1.179 123 0 thermal thermal e4:5f:01:7c:3f:34 1671222494.2805836 0.0005502700805664062 4.8 101858.0 MB 10.0 20221216202814_89526e7c-980b-4dbe-8257-c1c6944cdbd3 12/16/2022 15:28:18 1671222498 12/16/2022 15:28:14 2022-12-16 20:28:17.943010+00:00 26.9162 31.72 769.0 1.0 65535.0 99789.35 86.0
Time taken: 0.835 seconds, Fetched 10 row(s)
select uuid, ts, datetimestamp, co2, humidity, pressure, temperature
from local.ice_sink_thermal limit 10;
thrml_rwa_20221216203730 1671223055 2022-12-16 20:37:34.204439+00:00 783.0 32.22 99792.16 26.7613
thrml_qqs_20221216203735 1671223060 2022-12-16 20:37:39.013362+00:00 783.0 32.23 99791.39 26.78
thrml_szi_20221216203740 1671223064 2022-12-16 20:37:43.755430+00:00 784.0 32.14 99794.28 26.7934
thrml_rvb_20221216203744 1671223069 2022-12-16 20:37:48.563791+00:00 784.0 32.16 99796.71 26.8147
thrml_rto_20221216203749 1671223074 2022-12-16 20:37:53.373391+00:00 783.0 32.11 99796.0 26.8575
thrml_svv_20221216203754 1671223079 2022-12-16 20:37:58.184190+00:00 783.0 32.07 99791.21 26.8842
thrml_aov_20221216203759 1671223084 2022-12-16 20:38:02.991664+00:00 782.0 31.95 99794.04 26.9082
thrml_tzs_20221216203804 1671223088 2022-12-16 20:38:07.802992+00:00 782.0 32.02 99794.33 26.9322
thrml_fso_20221216203808 1671223093 2022-12-16 20:38:12.613666+00:00 783.0 32.05 99792.42 26.9589
thrml_czp_20221216203813 1671223098 2022-12-16 20:38:17.321001+00:00 783.0 31.98 99790.26 26.9803
Time taken: 0.898 seconds, Fetched 10 row(s)
References
- https://github.com/tspannhw/FLiP-Pi-DeltaLake-Thermal
- https://iceberg.apache.org/docs/latest/getting-started/
- https://github.com/streamnative/pulsar-io-lakehouse/blob/master/docs/lakehouse-sink.md
- https://streamnative.io/blog/release/2022-12-14-announcing-the-iceberg-sink-connector-for-apache-pulsar/
- https://hub.streamnative.io/connectors/lakehouse-sink/v2.10.1.12/
- https://github.com/tspannhw/FLiP-Pi-Thermal
- https://dzone.com/articles/pulsar-in-python-on-pi
- https://github.com/tabular-io/docker-spark-iceberg
- https://iceberg.apache.org/docs/latest/getting-started/
- https://stackoverflow.com/questions/73791829/delta-lake-sink-connector-for-apache-pulsar-with-minio-throws-java-lang-illegal
- https://thenewstack.io/apache-iceberg-a-different-table-design-for-big-data/
2022/2023 - Tim Spann - @PaaSDev