Mixing
PySpark creating tar file from s3 objects
I have small Spark job that collect files from s3, group them by key and save them to tar.
I'm running this job on large EMR cluster and i'm getting low performance.
The Input DataFrame size is ~10M-20M records.
Each id has ~1000 files.
When I use partitions = distinct_ids / 20, it work stable, each task take about 1 hour. stable but slow!
I tried to repartition by distinct_ids but its not working, start and then freeze.
Any idea how to improve performance of this job or other technic to group files into tar ?
def archive_files(f_list, str_date, source_bucket, destination_bucket):
f_cat = list()
for f in f_list:
# download file from s3
# add file to tar
f_cat.append = {list with file metadata}
# upload tar file to s3
return f_cat # list of dictionaries
if __name__ == '__main__':
with pyspark.SparkContext(appName="x") as sc:
sqlContext = SQLContext(sparkContext=sc)
df_cat = sqlContext.read.parquet(cat_path)
distinct_ids = df_cat.select("id").distinct().count()
res_cat_rdd = df_cat.rdd.groupBy(lambda x: x.id)
.repartition(distinct_ids)
.map(lambda r: archive_files(r, batch_name, SOURCE_BUCKET, DESTINATION_BUCKET))
.flatMap(lambda x: x)
res_cat_df = res_cat_rdd.toDF()
res_cat_df.write.mode('overwrite').parquet(cat_path_output)
pyspark amazon-emr
asked Nov 19 '18 at 21:35
alon_ralon_r
6315
add a comment |
I have small Spark job that collect files from s3, group them by key and save them to tar.
I'm running this job on large EMR cluster and i'm getting low performance.
The Input DataFrame size is ~10M-20M records.
Each id has ~1000 files.
When I use partitions = distinct_ids / 20, it work stable, each task take about 1 hour. stable but slow!
I tried to repartition by distinct_ids but its not working, start and then freeze.
Any idea how to improve performance of this job or other technic to group files into tar ?
def archive_files(f_list, str_date, source_bucket, destination_bucket):
f_cat = list()
for f in f_list:
# download file from s3
# add file to tar
f_cat.append = {list with file metadata}
# upload tar file to s3
return f_cat # list of dictionaries
if __name__ == '__main__':
with pyspark.SparkContext(appName="x") as sc:
sqlContext = SQLContext(sparkContext=sc)
df_cat = sqlContext.read.parquet(cat_path)
distinct_ids = df_cat.select("id").distinct().count()
res_cat_rdd = df_cat.rdd.groupBy(lambda x: x.id)
.repartition(distinct_ids)
.map(lambda r: archive_files(r, batch_name, SOURCE_BUCKET, DESTINATION_BUCKET))
.flatMap(lambda x: x)
res_cat_df = res_cat_rdd.toDF()
res_cat_df.write.mode('overwrite').parquet(cat_path_output)
pyspark amazon-emr
asked Nov 19 '18 at 21:35
alon_ralon_r
6315
add a comment |
I have small Spark job that collect files from s3, group them by key and save them to tar.
I'm running this job on large EMR cluster and i'm getting low performance.
The Input DataFrame size is ~10M-20M records.
Each id has ~1000 files.
When I use partitions = distinct_ids / 20, it work stable, each task take about 1 hour. stable but slow!
I tried to repartition by distinct_ids but its not working, start and then freeze.
Any idea how to improve performance of this job or other technic to group files into tar ?
def archive_files(f_list, str_date, source_bucket, destination_bucket):
f_cat = list()
for f in f_list:
# download file from s3
# add file to tar
f_cat.append = {list with file metadata}
# upload tar file to s3
return f_cat # list of dictionaries
if __name__ == '__main__':
with pyspark.SparkContext(appName="x") as sc:
sqlContext = SQLContext(sparkContext=sc)
df_cat = sqlContext.read.parquet(cat_path)
distinct_ids = df_cat.select("id").distinct().count()
res_cat_rdd = df_cat.rdd.groupBy(lambda x: x.id)
.repartition(distinct_ids)
.map(lambda r: archive_files(r, batch_name, SOURCE_BUCKET, DESTINATION_BUCKET))
.flatMap(lambda x: x)
res_cat_df = res_cat_rdd.toDF()
res_cat_df.write.mode('overwrite').parquet(cat_path_output)
pyspark amazon-emr
asked Nov 19 '18 at 21:35
alon_ralon_r
6315
I have small Spark job that collect files from s3, group them by key and save them to tar.
I'm running this job on large EMR cluster and i'm getting low performance.
The Input DataFrame size is ~10M-20M records.
Each id has ~1000 files.
When I use partitions = distinct_ids / 20, it work stable, each task take about 1 hour. stable but slow!
I tried to repartition by distinct_ids but its not working, start and then freeze.
Any idea how to improve performance of this job or other technic to group files into tar ?
def archive_files(f_list, str_date, source_bucket, destination_bucket):
f_cat = list()
for f in f_list:
# download file from s3
# add file to tar
f_cat.append = {list with file metadata}
# upload tar file to s3
return f_cat # list of dictionaries
if __name__ == '__main__':
with pyspark.SparkContext(appName="x") as sc:
sqlContext = SQLContext(sparkContext=sc)
df_cat = sqlContext.read.parquet(cat_path)
distinct_ids = df_cat.select("id").distinct().count()
res_cat_rdd = df_cat.rdd.groupBy(lambda x: x.id)
.repartition(distinct_ids)
.map(lambda r: archive_files(r, batch_name, SOURCE_BUCKET, DESTINATION_BUCKET))
.flatMap(lambda x: x)
res_cat_df = res_cat_rdd.toDF()
res_cat_df.write.mode('overwrite').parquet(cat_path_output)
pyspark amazon-emr
pyspark amazon-emr
asked Nov 19 '18 at 21:35
alon_ralon_r
6315
asked Nov 19 '18 at 21:35
alon_ralon_r
6315
asked Nov 19 '18 at 21:35
alon_ralon_r
6315
asked Nov 19 '18 at 21:35
alon_ralon_r
6315
asked Nov 19 '18 at 21:35
alon_ralon_r
6315
6315
add a comment |
add a comment |
0
active
oldest
votes
Your Answer
StackExchange.ifUsing("editor", function () {
StackExchange.using("externalEditor", function () {
StackExchange.using("snippets", function () {
StackExchange.snippets.init();
});
});
}, "code-snippets");
StackExchange.ready(function() {
var channelOptions = {
tags: "".split(" "),
id: "1"
};
initTagRenderer("".split(" "), "".split(" "), channelOptions);
StackExchange.using("externalEditor", function() {
// Have to fire editor after snippets, if snippets enabled
if (StackExchange.settings.snippets.snippetsEnabled) {
StackExchange.using("snippets", function() {
createEditor();
});
}
else {
createEditor();
}
});
function createEditor() {
StackExchange.prepareEditor({
heartbeatType: 'answer',
autoActivateHeartbeat: false,
convertImagesToLinks: true,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: 10,
bindNavPrevention: true,
postfix: "",
imageUploader: {
brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
allowUrls: true
},
onDemand: true,
discardSelector: ".discard-answer"
,immediatelyShowMarkdownHelp:true
});
}
});
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fstackoverflow.com%2fquestions%2f53382977%2fpyspark-creating-tar-file-from-s3-objects%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
0
active
oldest
votes
0
active
oldest
votes
active
oldest
votes
active
oldest
votes
Thanks for contributing an answer to Stack Overflow!
- Please be sure to answer the question. Provide details and share your research!
But avoid …
- Asking for help, clarification, or responding to other answers.
- Making statements based on opinion; back them up with references or personal experience.
To learn more, see our tips on writing great answers.
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fstackoverflow.com%2fquestions%2f53382977%2fpyspark-creating-tar-file-from-s3-objects%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
