Skip to content

bulk-cdk: source output performance improvements #44865

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.

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

Merged
merged 2 commits into from
Aug 28, 2024
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
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
Original file line number Diff line number Diff line change
@@ -1,6 +1,7 @@
/* Copyright (c) 2024 Airbyte, Inc., all rights reserved. */
package io.airbyte.cdk.output

import com.fasterxml.jackson.databind.SequenceWriter
import io.airbyte.cdk.util.Jsons
import io.airbyte.protocol.models.v0.AirbyteAnalyticsTraceMessage
import io.airbyte.protocol.models.v0.AirbyteCatalog
Expand Down Expand Up @@ -99,6 +100,7 @@ private class StdoutOutputConsumer : OutputConsumer {
override val emittedAt: Instant = Instant.now()

private val buffer = ByteArrayOutputStream()
private val sequenceWriter: SequenceWriter = Jsons.writer().writeValues(buffer)

override fun accept(airbyteMessage: AirbyteMessage) {
// This method effectively println's its JSON-serialized argument.
Expand All @@ -108,12 +110,12 @@ private class StdoutOutputConsumer : OutputConsumer {
// Other Airbyte message types are not buffered, instead they trigger an immediate flush.
// Such messages should not linger indefinitely in a buffer.
val isRecord: Boolean = airbyteMessage.type == AirbyteMessage.Type.RECORD
val json: ByteArray = Jsons.writeValueAsBytes(airbyteMessage)
synchronized(this) {
if (buffer.size() > 0) {
buffer.write('\n'.code)
}
buffer.writeBytes(json)
sequenceWriter.write(airbyteMessage)
sequenceWriter.flush()
if (!isRecord || buffer.size() >= BUFFER_MAX_SIZE) {
withLockFlush()
}
Expand Down
Original file line number Diff line number Diff line change
@@ -1,6 +1,7 @@
/* Copyright (c) 2024 Airbyte, Inc., all rights reserved. */
package io.airbyte.cdk.output

import io.airbyte.cdk.util.Jsons
import io.airbyte.protocol.models.v0.AirbyteCatalog
import io.airbyte.protocol.models.v0.AirbyteConnectionStatus
import io.airbyte.protocol.models.v0.AirbyteLogMessage
Expand Down Expand Up @@ -35,7 +36,10 @@ class BufferingOutputConsumer(
private val traces = mutableListOf<AirbyteTraceMessage>()
private val messages = mutableListOf<AirbyteMessage>()

override fun accept(m: AirbyteMessage) {
override fun accept(input: AirbyteMessage) {
// Deep copy the input, which may be reused and mutated later on.
val m: AirbyteMessage =
Jsons.readValue(Jsons.writeValueAsBytes(input), AirbyteMessage::class.java)
synchronized(this) {
messages.add(m)
when (m.type) {
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -335,7 +335,7 @@ data class TestCase(
)
for (actualState in actual!!) {
Assertions.assertTrue(
actualState in expected,
actualState.toString() in expected.map { it.toString() },
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

stringification is required because of the deep-copying added to BufferingOutputConsumer, this causes the hash values to change or something, failing the check for no good reason.

"$actualState should be in $expected",
)
}
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -6,6 +6,7 @@ import com.fasterxml.jackson.databind.node.ObjectNode
import io.airbyte.cdk.command.OpaqueStateValue
import io.airbyte.cdk.output.OutputConsumer
import io.airbyte.cdk.util.Jsons
import io.airbyte.protocol.models.v0.AirbyteMessage
import io.airbyte.protocol.models.v0.AirbyteRecordMessage
import java.util.concurrent.atomic.AtomicBoolean
import java.util.concurrent.atomic.AtomicLong
Expand Down Expand Up @@ -38,18 +39,25 @@ sealed class JdbcPartitionReader<P : JdbcPartition<*>>(
}

fun out(record: ObjectNode) {
val recordMessageData: ObjectNode = Jsons.objectNode()
for (fieldName in streamFieldNames) {
recordMessageData.set<JsonNode>(fieldName, record[fieldName] ?: Jsons.nullNode())
outData.set<JsonNode>(fieldName, record[fieldName] ?: Jsons.nullNode())
}
outputConsumer.accept(
AirbyteRecordMessage()
.withStream(stream.name)
.withNamespace(stream.namespace)
.withData(recordMessageData),
)
outputConsumer.accept(msg)
}

private val outData: ObjectNode = Jsons.objectNode()

private val msg =
AirbyteMessage()
.withType(AirbyteMessage.Type.RECORD)
.withRecord(
AirbyteRecordMessage()
.withEmittedAt(outputConsumer.emittedAt.toEpochMilli())
.withStream(stream.name)
.withNamespace(stream.namespace)
.withData(outData)
)

val streamFieldNames: List<String> = stream.fields.map { it.id }

override fun releaseResources() {
Expand All @@ -69,7 +77,11 @@ class JdbcNonResumablePartitionReader<P : JdbcPartition<*>>(
selectQuerier
.executeQuery(
q = partition.nonResumableQuery,
parameters = SelectQuerier.Parameters(streamState.fetchSize),
parameters =
SelectQuerier.Parameters(
reuseResultObject = true,
fetchSize = streamState.fetchSize
),
)
.use { result: SelectQuerier.Result ->
for (record in result) {
Expand Down Expand Up @@ -109,7 +121,8 @@ class JdbcResumablePartitionReader<P : JdbcSplittablePartition<*>>(
selectQuerier
.executeQuery(
q = partition.resumableQuery(limit),
parameters = SelectQuerier.Parameters(fetchSize),
parameters =
SelectQuerier.Parameters(reuseResultObject = true, fetchSize = fetchSize),
)
.use { result: SelectQuerier.Result ->
for (record in result) {
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -21,6 +21,9 @@ interface SelectQuerier {
): Result

data class Parameters(
/** When set, the [ObjectNode] in the [Result] is reused; take care with this! */
val reuseResultObject: Boolean = false,
/** JDBC fetchSize value. */
val fetchSize: Int? = null,
)

Expand All @@ -46,6 +49,7 @@ class JdbcSelectQuerier(
var conn: Connection? = null
var stmt: PreparedStatement? = null
var rs: ResultSet? = null
val reusable: ObjectNode? = Jsons.objectNode().takeIf { parameters.reuseResultObject }

init {
log.info { "Querying ${q.sql}" }
Expand Down Expand Up @@ -94,7 +98,7 @@ class JdbcSelectQuerier(
// necessary.
if (!hasNext()) throw NoSuchElementException()
// Read the current row in the ResultSet
val record: ObjectNode = Jsons.objectNode()
val record: ObjectNode = reusable ?: Jsons.objectNode()
var colIdx = 1
for (column in q.columns) {
log.debug { "Getting value #$colIdx for $column." }
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -47,7 +47,7 @@ class JdbcPartitionReaderTest {
)
),
),
SelectQuerier.Parameters(fetchSize = 2),
SelectQuerier.Parameters(reuseResultObject = true, fetchSize = 2),
"""{"id":1,"ts":"2024-08-01","msg":"hello"}""",
"""{"id":2,"ts":"2024-08-02","msg":"how"}""",
"""{"id":3,"ts":"2024-08-03","msg":"are"}""",
Expand Down Expand Up @@ -126,7 +126,7 @@ class JdbcPartitionReaderTest {
OrderBy(ts),
Limit(4),
),
SelectQuerier.Parameters(fetchSize = 2),
SelectQuerier.Parameters(reuseResultObject = true, fetchSize = 2),
"""{"id":1,"ts":"2024-08-01","msg":"hello"}""",
"""{"id":2,"ts":"2024-08-02","msg":"how"}""",
"""{"id":3,"ts":"2024-08-03","msg":"are"}""",
Expand Down
Original file line number Diff line number Diff line change
@@ -1,6 +1,7 @@
/* Copyright (c) 2024 Airbyte, Inc., all rights reserved. */
package io.airbyte.cdk.read

import com.fasterxml.jackson.databind.JsonNode
import com.fasterxml.jackson.databind.node.ObjectNode
import io.airbyte.cdk.discover.Field
import io.airbyte.cdk.h2.H2TestFixture
Expand Down Expand Up @@ -77,7 +78,7 @@ class JdbcSelectQuerierTest {

private fun runTest(
q: SelectQuery,
vararg expected: String,
vararg expectedJson: String,
) {
val configPojo: H2SourceConfigurationJsonObject =
H2SourceConfigurationJsonObject().apply {
Expand All @@ -86,7 +87,21 @@ class JdbcSelectQuerierTest {
}
val config: H2SourceConfiguration = H2SourceConfigurationFactory().make(configPojo)
val querier: SelectQuerier = JdbcSelectQuerier(JdbcConnectionFactory(config))
// Vanilla query
val expected: List<JsonNode> = expectedJson.map(Jsons::readTree)
val actual: List<ObjectNode> = querier.executeQuery(q).use { it.asSequence().toList() }
Assertions.assertIterableEquals(expected.toList().map(Jsons::readTree), actual)
Assertions.assertIterableEquals(expected, actual)
// Query with reuseResultObject = true
querier.executeQuery(q, SelectQuerier.Parameters(reuseResultObject = true)).use {
var i = 0
var previous: ObjectNode? = null
for (record in it) {
if (i > 0) {
Assertions.assertTrue(previous === record)
}
Assertions.assertEquals(expected[i++], record)
previous = record
}
}
}
}
Loading