Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[FSTORE-1640] Delta delta streamer cache dataframes #432

Open
wants to merge 4 commits into
base: main
Choose a base branch
from
Open

[FSTORE-1640] Delta delta streamer cache dataframes #432

Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
840824e
init
bubriks Dec 9, 2024
be76f78
small change
bubriks Dec 9, 2024
0d2dd74
add endingOffsets to delta delta streamer
bubriks Dec 9, 2024
a19f2b1
small fix
bubriks Dec 9, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
25 changes: 18 additions & 7 deletions utils/python/hsfs_utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -273,13 +273,13 @@ def offline_fg_materialization(spark: SparkSession, job_conf: Dict[Any, Any], in
entity.feature_store_id, {}, engine="spark"
)

# get offsets
# get starting offsets
offset_location = entity.prepare_spark_location() + "/kafka_offsets"
try:
if initial_check_point_string:
offset_string = json.dumps(_build_starting_offsets(initial_check_point_string))
starting_offset_string = json.dumps(_build_starting_offsets(initial_check_point_string))
else:
offset_string = spark.read.json(offset_location).toJSON().first()
starting_offset_string = spark.read.json(offset_location).toJSON().first()
except Exception as e:
print(f"An unexpected error occurred: {e}")
# if all else fails read from the beggining
Expand All @@ -289,15 +289,26 @@ def offline_fg_materialization(spark: SparkSession, job_conf: Dict[Any, Any], in
offline_write_options={},
high=False,
)
offset_string = json.dumps(_build_starting_offsets(initial_check_point_string))
print(f"startingOffsets: {offset_string}")
starting_offset_string = json.dumps(_build_starting_offsets(initial_check_point_string))
print(f"startingOffsets: {starting_offset_string}")

# get ending offsets
ending_offset_string = kafka_engine.kafka_get_offsets(
topic_name=entity._online_topic_name,
feature_store_id=entity.feature_store_id,
offline_write_options={},
high=True,
)
ending_offset_string = json.dumps(_build_starting_offsets(ending_offset_string))
print(f"endingOffsets: {ending_offset_string}")

# read kafka topic
df = (
spark.read.format("kafka")
.options(**read_options)
.option("subscribe", entity._online_topic_name)
.option("startingOffsets", offset_string)
.option("startingOffsets", starting_offset_string)
.option("endingOffsets", ending_offset_string)
.option("includeHeaders", "true")
.option("failOnDataLoss", "false")
.load()
Expand All @@ -320,7 +331,7 @@ def offline_fg_materialization(spark: SparkSession, job_conf: Dict[Any, Any], in

# update offsets
df_offsets = (df if limit > filtered_df.count() else filtered_df).groupBy('partition').agg(max('offset').alias('offset')).collect()
offset_dict = json.loads(offset_string)
offset_dict = json.loads(starting_offset_string)
for offset_row in df_offsets:
offset_dict[f"{entity._online_topic_name}"][f"{offset_row.partition}"] = offset_row.offset + 1

Expand Down
Loading