Re: Correctness bug on Shuffle+Repartition scenario
Hi,
I am using Spark 2.4.4 standalone mode.
On Mon, Jan 18, 2021 at 4:26 AM Sean Owen wrote:
> Hm, FWIW I can't reproduce that on Spark 3.0.1. What version are you using?
>
> On Sun, Jan 17, 2021 at 6:22 AM Shiao-An Yuan
> wrote:
>
>> Hi folks,
>>
>> I finally found the root cause of this issue.
>> It can be easily reproduced by the following code.
>> We ran it on a standalone mode 4 cores * 4 instances (total 16 cores)
>> environment.
>>
>> ```
>> import org.apache.spark.TaskContext
>> import scala.sys.process._
>> import org.apache.spark.sql.functions._
>> import com.google.common.hash.Hashing
>> val murmur3 = Hashing.murmur3_32()
>>
>> // create a Dataset with the cardinality of the second element equals
>> 5.
>> val ds = spark.range(0, 10, 1, 130).map(i =>
>> (murmur3.hashLong(i).asInt(), i/2))
>>
>> ds.groupByKey(_._2)
>> .agg(first($"_1").as[Long])
>> .repartition(200)
>> .map { x =>
>> if (TaskContext.get.attemptNumber == 0 && TaskContext.get.partitionId
>> == 100 && TaskContext.get.stageAttemptNumber == 0) {
>> throw new Exception("pkill -f CoarseGrainedExecutorBackend".!!)
>> }
>> x
>> }
>> .map(_._2).distinct().count() // the correct result is 5, but we
>> always got fewer number
>> ```
>>
>> The problem here is SPARK-23207 use sorting to make
>> RoundRobinPartitioning always generate the same distribution,
>> but the UDAF `first` may return non-deterministic results and caused the
>> sorting result non-deterministic.
>> Therefore, the first stage and the retry stage might have different
>> distribution and cause duplications and loss.
>>
>> Thanks,
>> Shiao-An Yuan
>>
>> On Tue, Dec 29, 2020 at 10:00 PM Shiao-An Yuan
>> wrote:
>>
>>> Hi folks,
>>>
>>> We recently identified a data correctness issue in our pipeline.
>>>
>>> The data processing flow is as follows:
>>> 1. read the current snapshot (provide empty if it doesn't exist yet)
>>> 2. read unprocessed new data
>>> 3. union them and do a `reduceByKey` operation
>>> 4. output a new version of the snapshot
>>> 5. repeat step 1~4
>>>
>>> The simplified version of code:
>>> ```
>>> // schema
>>> case class Log(pkey: Array[Byte], a: String, b: Int, /* 100+ columns */)
>>>
>>> // function for reduce
>>> def merge(left: Log, right: Log): Log = {
>>> Log(pkey = left.pkey
>>> a= if (left.a!=null) left.a else right.a,
>>> b= if (left.a!=null) left.b else right.b,
>>> ...
>>> )
>>> }
>>>
>>> // a very large parquet file (>10G, 200 partitions)
>>> val currentSnapshot = spark.read.schema(schema).parquet(...).as[Log]
>>>
>>> // multiple small parquet files
>>> val newAddedLogs = spark.read.schema(schema).parquet(...).as[Log]
>>>
>>> val newSnapshot = currentSnapshot.union(newAddedLog)
>>> .groupByKey(new String(pkey)) // generate key
>>> .reduceGroups(_.merge(_))//
>>> spark.sql.shuffle.partitions=200
>>> .map(_._2) // drop key
>>>
>>> newSnapshot
>>> .repartition(60) // (1)
>>> .write.parquet(newPath)
>>> ```
>>>
>>> The issue we have is that some data were duplicated or lost, and the
>>> amount of
>>> duplicated and loss data are similar.
>>>
>>> We also noticed that this situation only happens if some instances got
>>> preempted. Spark will retry the stage, so some of the partitioned files
>>> are
>>> generated at the 1st time, and other files are generated at the
>>> 2nd(retry) time.
>>> Moreover, those duplicated logs will be duplicated exactly twice and
>>> located in
>>> both batches (one in the first batch; and one in the second batch).
>>>
>>> The input/output files are parquet on GCS. The Spark version is 2.4.4
>>> with
>>> standalone deployment. Workers running on GCP preemptible instances and
>>> they
>>> being preempted very frequently.
>>>
>>> The pipeline is running in a single long-running process with
>>> multi-threads,
>>> each snapshot represent an "hour" of data, and we do the
>>> "read-reduce-write" operations
>>> on multiple snapshots(hours) simultaneously. We pretty sure the same
>>> snapshot(hour) never process parallelly and the output path always
>>> generated with a timestamp, so those jobs shouldn't affect each other.
>>>
>>> After changing the line (1) to `coalesce` or `repartition(100, $"pkey")`
>>> the issue
>>> was gone, but I believe there is still a correctness bug that hasn't
>>> been reported yet.
>>>
>>> We have tried to reproduce this bug on a smaller scale but haven't
>>> succeeded yet. I
>>> have read SPARK-23207 and SPARK-28699, but couldn't found the bug.
>>> Since this case is DataSet, I believe it is unrelated to SPARK-24243.
>>>
>>> Can anyone give me some advice about the following tasks?
>>> Thanks in advance.
>>>
>>> Shiao-An Yuan
>>>
>>
Re: Correctness bug on Shuffle+Repartition scenario
Hi folks,
I finally found the root cause of this issue.
It can be easily reproduced by the following code.
We ran it on a standalone mode 4 cores * 4 instances (total 16 cores)
environment.
```
import org.apache.spark.TaskContext
import scala.sys.process._
import org.apache.spark.sql.functions._
import com.google.common.hash.Hashing
val murmur3 = Hashing.murmur3_32()
// create a Dataset with the cardinality of the second element equals 5.
val ds = spark.range(0, 10, 1, 130).map(i =>
(murmur3.hashLong(i).asInt(), i/2))
ds.groupByKey(_._2)
.agg(first($"_1").as[Long])
.repartition(200)
.map { x =>
if (TaskContext.get.attemptNumber == 0 && TaskContext.get.partitionId
== 100 && TaskContext.get.stageAttemptNumber == 0) {
throw new Exception("pkill -f CoarseGrainedExecutorBackend".!!)
}
x
}
.map(_._2).distinct().count() // the correct result is 5, but we
always got fewer number
```
The problem here is SPARK-23207 use sorting to make RoundRobinPartitioning
always generate the same distribution,
but the UDAF `first` may return non-deterministic results and caused the
sorting result non-deterministic.
Therefore, the first stage and the retry stage might have different
distribution and cause duplications and loss.
Thanks,
Shiao-An Yuan
On Tue, Dec 29, 2020 at 10:00 PM Shiao-An Yuan
wrote:
> Hi folks,
>
> We recently identified a data correctness issue in our pipeline.
>
> The data processing flow is as follows:
> 1. read the current snapshot (provide empty if it doesn't exist yet)
> 2. read unprocessed new data
> 3. union them and do a `reduceByKey` operation
> 4. output a new version of the snapshot
> 5. repeat step 1~4
>
> The simplified version of code:
> ```
> // schema
> case class Log(pkey: Array[Byte], a: String, b: Int, /* 100+ columns */)
>
> // function for reduce
> def merge(left: Log, right: Log): Log = {
> Log(pkey = left.pkey
> a= if (left.a!=null) left.a else right.a,
> b= if (left.a!=null) left.b else right.b,
> ...
> )
> }
>
> // a very large parquet file (>10G, 200 partitions)
> val currentSnapshot = spark.read.schema(schema).parquet(...).as[Log]
>
> // multiple small parquet files
> val newAddedLogs = spark.read.schema(schema).parquet(...).as[Log]
>
> val newSnapshot = currentSnapshot.union(newAddedLog)
> .groupByKey(new String(pkey)) // generate key
> .reduceGroups(_.merge(_))//
> spark.sql.shuffle.partitions=200
> .map(_._2) // drop key
>
> newSnapshot
> .repartition(60) // (1)
> .write.parquet(newPath)
> ```
>
> The issue we have is that some data were duplicated or lost, and the
> amount of
> duplicated and loss data are similar.
>
> We also noticed that this situation only happens if some instances got
> preempted. Spark will retry the stage, so some of the partitioned files are
> generated at the 1st time, and other files are generated at the 2nd(retry)
> time.
> Moreover, those duplicated logs will be duplicated exactly twice and
> located in
> both batches (one in the first batch; and one in the second batch).
>
> The input/output files are parquet on GCS. The Spark version is 2.4.4 with
> standalone deployment. Workers running on GCP preemptible instances and
> they
> being preempted very frequently.
>
> The pipeline is running in a single long-running process with
> multi-threads,
> each snapshot represent an "hour" of data, and we do the
> "read-reduce-write" operations
> on multiple snapshots(hours) simultaneously. We pretty sure the same
> snapshot(hour) never process parallelly and the output path always
> generated with a timestamp, so those jobs shouldn't affect each other.
>
> After changing the line (1) to `coalesce` or `repartition(100, $"pkey")`
> the issue
> was gone, but I believe there is still a correctness bug that hasn't been
> reported yet.
>
> We have tried to reproduce this bug on a smaller scale but haven't
> succeeded yet. I
> have read SPARK-23207 and SPARK-28699, but couldn't found the bug.
> Since this case is DataSet, I believe it is unrelated to SPARK-24243.
>
> Can anyone give me some advice about the following tasks?
> Thanks in advance.
>
> Shiao-An Yuan
>
Re: Correctness bug on Shuffle+Repartition scenario
Hi Sean,
Sorry, I didn't describe it clearly. The column "pkey" is like a "Primary
Key" and I do "reduce by key" on this column, so the "amount of rows"
should always equal to the "cardinality of pkey".
When I said data get duplicated & lost, I mean duplicated "pkey" exists in
the output file (after "reduce by key") and some "pkey" missing.
Since it only happens when executors being preempted, I believe this is a
bug (nondeterministic shuffle) that SPARK-23207 trying to solve.
Thanks,
Shiao-An Yuan
On Tue, Dec 29, 2020 at 10:53 PM Sean Owen wrote:
> Total guess here, but your key is a case class. It does define hashCode
> and equals for you, but, you have an array as one of the members. Array
> equality is by reference, so, two arrays of the same elements are not
> equal. You may have to define hashCode and equals manually to make them
> correct.
>
> On Tue, Dec 29, 2020 at 8:01 AM Shiao-An Yuan
> wrote:
>
>> Hi folks,
>>
>> We recently identified a data correctness issue in our pipeline.
>>
>> The data processing flow is as follows:
>> 1. read the current snapshot (provide empty if it doesn't exist yet)
>> 2. read unprocessed new data
>> 3. union them and do a `reduceByKey` operation
>> 4. output a new version of the snapshot
>> 5. repeat step 1~4
>>
>> The simplified version of code:
>> ```
>> // schema
>> case class Log(pkey: Array[Byte], a: String, b: Int, /* 100+ columns */)
>>
>> // function for reduce
>> def merge(left: Log, right: Log): Log = {
>> Log(pkey = left.pkey
>> a= if (left.a!=null) left.a else right.a,
>> b= if (left.a!=null) left.b else right.b,
>> ...
>> )
>> }
>>
>> // a very large parquet file (>10G, 200 partitions)
>> val currentSnapshot = spark.read.schema(schema).parquet(...).as[Log]
>>
>> // multiple small parquet files
>> val newAddedLogs = spark.read.schema(schema).parquet(...).as[Log]
>>
>> val newSnapshot = currentSnapshot.union(newAddedLog)
>> .groupByKey(new String(pkey)) // generate key
>> .reduceGroups(_.merge(_))//
>> spark.sql.shuffle.partitions=200
>> .map(_._2) // drop key
>>
>> newSnapshot
>> .repartition(60) // (1)
>> .write.parquet(newPath)
>> ```
>>
>> The issue we have is that some data were duplicated or lost, and the
>> amount of
>> duplicated and loss data are similar.
>>
>> We also noticed that this situation only happens if some instances got
>> preempted. Spark will retry the stage, so some of the partitioned files
>> are
>> generated at the 1st time, and other files are generated at the
>> 2nd(retry) time.
>> Moreover, those duplicated logs will be duplicated exactly twice and
>> located in
>> both batches (one in the first batch; and one in the second batch).
>>
>> The input/output files are parquet on GCS. The Spark version is 2.4.4 with
>> standalone deployment. Workers running on GCP preemptible instances and
>> they
>> being preempted very frequently.
>>
>> The pipeline is running in a single long-running process with
>> multi-threads,
>> each snapshot represent an "hour" of data, and we do the
>> "read-reduce-write" operations
>> on multiple snapshots(hours) simultaneously. We pretty sure the same
>> snapshot(hour) never process parallelly and the output path always
>> generated with a timestamp, so those jobs shouldn't affect each other.
>>
>> After changing the line (1) to `coalesce` or `repartition(100, $"pkey")`
>> the issue
>> was gone, but I believe there is still a correctness bug that hasn't been
>> reported yet.
>>
>> We have tried to reproduce this bug on a smaller scale but haven't
>> succeeded yet. I
>> have read SPARK-23207 and SPARK-28699, but couldn't found the bug.
>> Since this case is DataSet, I believe it is unrelated to SPARK-24243.
>>
>> Can anyone give me some advice about the following tasks?
>> Thanks in advance.
>>
>> Shiao-An Yuan
>>
>
Correctness bug on Shuffle+Repartition scenario
Hi folks,
We recently identified a data correctness issue in our pipeline.
The data processing flow is as follows:
1. read the current snapshot (provide empty if it doesn't exist yet)
2. read unprocessed new data
3. union them and do a `reduceByKey` operation
4. output a new version of the snapshot
5. repeat step 1~4
The simplified version of code:
```
// schema
case class Log(pkey: Array[Byte], a: String, b: Int, /* 100+ columns */)
// function for reduce
def merge(left: Log, right: Log): Log = {
Log(pkey = left.pkey
a= if (left.a!=null) left.a else right.a,
b= if (left.a!=null) left.b else right.b,
...
)
}
// a very large parquet file (>10G, 200 partitions)
val currentSnapshot = spark.read.schema(schema).parquet(...).as[Log]
// multiple small parquet files
val newAddedLogs = spark.read.schema(schema).parquet(...).as[Log]
val newSnapshot = currentSnapshot.union(newAddedLog)
.groupByKey(new String(pkey)) // generate key
.reduceGroups(_.merge(_))//
spark.sql.shuffle.partitions=200
.map(_._2) // drop key
newSnapshot
.repartition(60) // (1)
.write.parquet(newPath)
```
The issue we have is that some data were duplicated or lost, and the amount
of
duplicated and loss data are similar.
We also noticed that this situation only happens if some instances got
preempted. Spark will retry the stage, so some of the partitioned files are
generated at the 1st time, and other files are generated at the 2nd(retry)
time.
Moreover, those duplicated logs will be duplicated exactly twice and
located in
both batches (one in the first batch; and one in the second batch).
The input/output files are parquet on GCS. The Spark version is 2.4.4 with
standalone deployment. Workers running on GCP preemptible instances and they
being preempted very frequently.
The pipeline is running in a single long-running process with multi-threads,
each snapshot represent an "hour" of data, and we do the
"read-reduce-write" operations
on multiple snapshots(hours) simultaneously. We pretty sure the same
snapshot(hour) never process parallelly and the output path always
generated with a timestamp, so those jobs shouldn't affect each other.
After changing the line (1) to `coalesce` or `repartition(100, $"pkey")`
the issue
was gone, but I believe there is still a correctness bug that hasn't been
reported yet.
We have tried to reproduce this bug on a smaller scale but haven't
succeeded yet. I
have read SPARK-23207 and SPARK-28699, but couldn't found the bug.
Since this case is DataSet, I believe it is unrelated to SPARK-24243.
Can anyone give me some advice about the following tasks?
Thanks in advance.
Shiao-An Yuan
