HADOOP-19066. S3A: AWS SDK V2 - Enabling FIPS should be allowed with central endpoint #6539

virajjasani · 2024-02-08T07:21:26Z

Jira: HADOOP-19066

…central endpoint

hadoop-yetus · 2024-02-08T09:57:25Z

🎊 +1 overall

Vote	Subsystem	Runtime	Logfile	Comment
+0 🆗	reexec	17m 41s		Docker mode activated.
			_ Prechecks _
+1 💚	dupname	0m 0s		No case conflicting files found.
+0 🆗	codespell	0m 0s		codespell was not available.
+0 🆗	detsecrets	0m 0s		detect-secrets was not available.
+0 🆗	markdownlint	0m 0s		markdownlint was not available.
+1 💚	@author	0m 0s		The patch does not contain any @author tags.
+1 💚	test4tests	0m 0s		The patch appears to include 1 new or modified test files.
			_ trunk Compile Tests _
+1 💚	mvninstall	46m 25s		trunk passed
+1 💚	compile	0m 41s		trunk passed with JDK Ubuntu-11.0.21+9-post-Ubuntu-0ubuntu120.04
+1 💚	compile	0m 31s		trunk passed with JDK Private Build-1.8.0_392-8u392-ga-1~20.04-b08
+1 💚	checkstyle	0m 29s		trunk passed
+1 💚	mvnsite	0m 40s		trunk passed
+1 💚	javadoc	0m 24s		trunk passed with JDK Ubuntu-11.0.21+9-post-Ubuntu-0ubuntu120.04
+1 💚	javadoc	0m 31s		trunk passed with JDK Private Build-1.8.0_392-8u392-ga-1~20.04-b08
+1 💚	spotbugs	1m 9s		trunk passed
+1 💚	shadedclient	37m 30s		branch has no errors when building and testing our client artifacts.
			_ Patch Compile Tests _
+1 💚	mvninstall	0m 29s		the patch passed
+1 💚	compile	0m 33s		the patch passed with JDK Ubuntu-11.0.21+9-post-Ubuntu-0ubuntu120.04
+1 💚	javac	0m 33s		the patch passed
+1 💚	compile	0m 25s		the patch passed with JDK Private Build-1.8.0_392-8u392-ga-1~20.04-b08
+1 💚	javac	0m 25s		the patch passed
+1 💚	blanks	0m 0s		The patch has no blanks issues.
+1 💚	checkstyle	0m 20s		the patch passed
+1 💚	mvnsite	0m 30s		the patch passed
+1 💚	javadoc	0m 15s		the patch passed with JDK Ubuntu-11.0.21+9-post-Ubuntu-0ubuntu120.04
+1 💚	javadoc	0m 24s		the patch passed with JDK Private Build-1.8.0_392-8u392-ga-1~20.04-b08
+1 💚	spotbugs	1m 6s		the patch passed
+1 💚	shadedclient	37m 18s		patch has no errors when building and testing our client artifacts.
			_ Other Tests _
+1 💚	unit	2m 53s		hadoop-aws in the patch passed.
+1 💚	asflicense	0m 35s		The patch does not generate ASF License warnings.
		154m 41s

Subsystem	Report/Notes
Docker	ClientAPI=1.44 ServerAPI=1.44 base: https://ci-hadoop.apache.org/job/hadoop-multibranch/job/PR-6539/1/artifact/out/Dockerfile
GITHUB PR	#6539
Optional Tests	dupname asflicense compile javac javadoc mvninstall mvnsite unit shadedclient spotbugs checkstyle codespell detsecrets markdownlint
uname	Linux e0ea602ff831 5.15.0-88-generic #98-Ubuntu SMP Mon Oct 2 15:18:56 UTC 2023 x86_64 x86_64 x86_64 GNU/Linux
Build tool	maven
Personality	dev-support/bin/hadoop.sh
git revision	trunk / `fa91891`
Default Java	Private Build-1.8.0_392-8u392-ga-1~20.04-b08
Multi-JDK versions	/usr/lib/jvm/java-11-openjdk-amd64:Ubuntu-11.0.21+9-post-Ubuntu-0ubuntu120.04 /usr/lib/jvm/java-8-openjdk-amd64:Private Build-1.8.0_392-8u392-ga-1~20.04-b08
Test Results	https://ci-hadoop.apache.org/job/hadoop-multibranch/job/PR-6539/1/testReport/
Max. process+thread count	580 (vs. ulimit of 5500)
modules	C: hadoop-tools/hadoop-aws U: hadoop-tools/hadoop-aws
Console output	https://ci-hadoop.apache.org/job/hadoop-multibranch/job/PR-6539/1/console
versions	git=2.25.1 maven=3.6.3 spotbugs=4.2.2
Powered by	Apache Yetus 0.14.0 https://yetus.apache.org

This message was automatically generated.

virajjasani · 2024-02-09T01:01:50Z

Tested against us-west-2 bucket with endpoints: s3.amazonaws.com and s3-us-west-2.amazonaws.com:

mvn clean verify -Dparallel-tests -DtestsThreadCount=8 -Dscale -Dprefetch

virajjasani · 2024-03-08T07:50:33Z

@ahmarsuhail @mukund-thakur could you please review this PR?

steveloughran

+1; lets merge and see if any regressions surface

steveloughran · 2024-03-12T18:50:17Z

(testing cherrypick; if all is good will merge to 3.4.x)

steveloughran · 2024-03-12T19:19:31Z

not good on branch-3.4; we need a followup i'm afraid. leaving in trunk rather than reverting for now as the other tests all seem happy.

virajjasani · 2024-03-12T19:20:47Z

I will re-run the test suite and followup.

steveloughran · 2024-03-12T19:30:39Z

looking at my current settings I've set endpoint to london but the region is unset; making sure that the classic binding mechanism still works.


  <property>
    <name>fs.s3a.bucket.stevel-london.endpoint</name>
    <value>${london.endpoint}</value>
  </property>

  <property>
    <name>X.fs.s3a.bucket.stevel-london.endpoint.region</name>
    <value>${london.region}</value>
  </property>

virajjasani · 2024-03-12T19:49:37Z

rebasing and rebuilding both trunk and branch-3.4 before re-running the tests.

virajjasani · 2024-03-12T20:19:56Z

Something seems odd. This test overrides endpoint/region configs so setting any endpoint/region should have made no difference:

  @Test
  public void testCentralEndpointAndNullRegionFipsWithCRUD() throws Throwable {
    describe("Access the test bucket using central endpoint and"
        + " null region and fips enabled, perform file system CRUD operations");
    final Configuration conf = getConfiguration();

    final Configuration newConf = new Configuration(conf);

    removeBaseAndBucketOverrides(
        newConf,
        ENDPOINT,
        AWS_REGION,
        FIPS_ENDPOINT);

    newConf.set(ENDPOINT, CENTRAL_ENDPOINT);
    newConf.setBoolean(FIPS_ENDPOINT, true);

    newFS = new S3AFileSystem();
    newFS.initialize(getFileSystem().getUri(), newConf);

    assertOpsUsingNewFs();
  }

I tested using these settings and there is no difference in behaviour because the test overrides base and bucket configs for endpoint/region.

I tried:

endpoint: us-west-2, region: unset
endpoint: central, region: unset
endpoint: unset, region: unset
endpoint: us-west-2, region: us-west-2

From the stacktrace from Jira:

[ERROR] Tests run: 18, Failures: 0, Errors: 1, Skipped: 1, Time elapsed: 56.26 s <<< FAILURE! - in org.apache.hadoop.fs.s3a.ITestS3AEndpointRegion
[ERROR] testCentralEndpointAndNullRegionFipsWithCRUD(org.apache.hadoop.fs.s3a.ITestS3AEndpointRegion)  Time elapsed: 4.821 s  <<< ERROR!
java.net.UnknownHostException: getFileStatus on s3a://stevel-london/test/testCentralEndpointAndNullRegionFipsWithCRUD/srcdir: software.amazon.awssdk.core.exception.SdkClientException: Received an UnknownHostException when attempting to interact with a service. See cause for the exact endpoint that is failing to resolve. If this is happening on an endpoint that previously worked, there may be a network connectivity issue or your DNS cache could be storing endpoints for too long.:    software.amazon.awssdk.core.exception.SdkClientException: Received an UnknownHostException when attempting to interact with a service. See cause for the exact endpoint that is failing to resolve. If this is happening on an endpoint that previously worked, there may be a network connectivity issue or your DNS cache could be storing endpoints for too long.: stevel-london.s3-fips.eu-west-2.amazonaws.com
	at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)
	at sun.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructorAccessorImpl.java:62)
	at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:45)
	at java.lang.reflect.Constructor.newInstance(Constructor.java:423)
	at org.apache.hadoop.fs.s3a.impl.ErrorTranslation.wrapWithInnerIOE(ErrorTranslation.java:182)
	at org.apache.hadoop.fs.s3a.impl.ErrorTranslation.maybeExtractIOException(ErrorTranslation.java:152)
	at org.apache.hadoop.fs.s3a.S3AUtils.translateException(S3AUtils.java:207)
	at org.apache.hadoop.fs.s3a.S3AUtils.translateException(S3AUtils.java:155)
	at org.apache.hadoop.fs.s3a.S3AFileSystem.s3GetFileStatus(S3AFileSystem.java:4066)
	at org.apache.hadoop.fs.s3a.S3AFileSystem.innerGetFileStatus(S3AFileSystem.java:3922)
	at org.apache.hadoop.fs.s3a.S3AFileSystem$MkdirOperationCallbacksImpl.probePathStatus(S3AFileSystem.java:3794)
	at org.apache.hadoop.fs.s3a.impl.MkdirOperation.probePathStatusOrNull(MkdirOperation.java:173)
	at org.apache.hadoop.fs.s3a.impl.MkdirOperation.getPathStatusExpectingDir(MkdirOperation.java:194)
	at org.apache.hadoop.fs.s3a.impl.MkdirOperation.execute(MkdirOperation.java:108)
	at org.apache.hadoop.fs.s3a.impl.MkdirOperation.execute(MkdirOperation.java:57)
	at org.apache.hadoop.fs.s3a.impl.ExecutingStoreOperation.apply(ExecutingStoreOperation.java:76)
	at org.apache.hadoop.fs.statistics.impl.IOStatisticsBinding.invokeTrackingDuration(IOStatisticsBinding.java:547)
	at org.apache.hadoop.fs.statistics.impl.IOStatisticsBinding.lambda$trackDurationOfOperation$5(IOStatisticsBinding.java:528)
	at org.apache.hadoop.fs.statistics.impl.IOStatisticsBinding.trackDuration(IOStatisticsBinding.java:449)
	at org.apache.hadoop.fs.s3a.S3AFileSystem.trackDurationAndSpan(S3AFileSystem.java:2707)
	at org.apache.hadoop.fs.s3a.S3AFileSystem.trackDurationAndSpan(S3AFileSystem.java:2726)
	at org.apache.hadoop.fs.s3a.S3AFileSystem.mkdirs(S3AFileSystem.java:3766)
	at org.apache.hadoop.fs.FileSystem.mkdirs(FileSystem.java:2494)
	at org.apache.hadoop.fs.s3a.ITestS3AEndpointRegion.assertOpsUsingNewFs(ITestS3AEndpointRegion.java:461)
	at org.apache.hadoop.fs.s3a.ITestS3AEndpointRegion.testCentralEndpointAndNullRegionFipsWithCRUD(ITestS3AEndpointRegion.java:454)

Here, we set:

    removeBaseAndBucketOverrides(
        newConf,
        ENDPOINT,
        AWS_REGION,
        FIPS_ENDPOINT);

    newConf.set(ENDPOINT, CENTRAL_ENDPOINT);
    newConf.setBoolean(FIPS_ENDPOINT, true);

    newFS = new S3AFileSystem();
    newFS.initialize(getFileSystem().getUri(), newConf);

How could stacktrace show different region than us-east-2 when the test overrides endpoint to central and removes region?

virajjasani · 2024-03-12T20:35:59Z

Just created a bucket in london and now i can reproduce the failure, checking.
Fails for london bucket, even if central endpoint is used.

virajjasani · 2024-03-12T21:10:38Z

Issue seems with FIPS cases.

FIPS enabled and

bucket created on oregon, s3 client configured with us-east-2 region with cross-region access enabled and no endpoint override: things look good
bucket created on london, s3 client configured with us-east-2 region with cross-region access enabled and no endpoint override: fails with

Caused by: software.amazon.awssdk.core.exception.SdkClientException: Received an UnknownHostException when attempting to interact with a service. See cause for the exact endpoint that is failing to resolve. If this is happening on an endpoint that previously worked, there may be a network connectivity issue or your DNS cache could be storing endpoints for too long.

bucket created on paris, s3 client configured with us-east-2 region with cross-region access enabled and no endpoint override: fails with

Caused by: software.amazon.awssdk.core.exception.SdkClientException: Received an UnknownHostException when attempting to interact with a service. See cause for the exact endpoint that is failing to resolve. If this is happening on an endpoint that previously worked, there may be a network connectivity issue or your DNS cache could be storing endpoints for too long.

All of above cases pass if FIPS is disabled.

will create an SDK issue soon.

virajjasani · 2024-03-13T05:06:59Z

Oh wait, FIPS is only for US and Canada endpoints. The above error is legit.

Let me provide an addendum to ignore the test if non-US or Canada endpoints are used.

virajjasani · 2024-03-13T05:59:00Z

Addendum PR: #6624

steveloughran · 2024-03-13T12:35:27Z

thanks. always good to have a broad set of test configs amongst other devs, especially now there are things like s3 express. Milan and Jakarta and any other Post 2019 region are also trouble as central DNS doesn't resolve bucket names as stevel-milan.s3.amazonaws.com

…central endpoint (#6539) Includes HADOOP-19066. Run FIPS test for valid bucket locations (ADDENDUM) (#6624) FIPS is only supported in north america AWS regions; relevant tests in ITestS3AEndpointRegion are skipped for buckets with different endpoints/regions. Contributed by Viraj Jasani

virajjasani · 2024-03-14T00:26:01Z

Post 2019 region are also trouble as central DNS doesn't resolve bucket names

indeed, that is also problematic.

steveloughran · 2024-03-15T12:31:04Z

FWIW a real problem is that the v2 sdk retries on unknown host exception until timeout, and that inner exception is lost.

Add an option to disable retries on UnknownHostException aws/aws-sdk-java-v2#4739
ApiCallTimeoutException fails to include exception of last failure; loses root cause and so breaks recovery logic of applications. aws/aws-sdk-java-v2#4738

I see there may be ways in the aws sdk to restrict more exceptions to retry; probably merits investigation.

HADOOP-19066. S3A: AWS SDK V2 - Enabling FIPS should be allowed with …

fa91891

…central endpoint

github-actions bot added trunk TOOLS AWS labels Feb 8, 2024

steveloughran approved these changes Mar 12, 2024

View reviewed changes

steveloughran merged commit 44c14ed into apache:trunk Mar 12, 2024
4 checks passed

virajjasani mentioned this pull request Mar 13, 2024

HADOOP-19066. Run FIPS test for valid bucket locations (ADDENDUM) #6624

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

HADOOP-19066. S3A: AWS SDK V2 - Enabling FIPS should be allowed with central endpoint #6539

HADOOP-19066. S3A: AWS SDK V2 - Enabling FIPS should be allowed with central endpoint #6539

virajjasani commented Feb 8, 2024

hadoop-yetus commented Feb 8, 2024

virajjasani commented Feb 9, 2024

virajjasani commented Mar 8, 2024

steveloughran left a comment

steveloughran commented Mar 12, 2024

steveloughran commented Mar 12, 2024

virajjasani commented Mar 12, 2024

steveloughran commented Mar 12, 2024 •

edited

Loading

virajjasani commented Mar 12, 2024 •

edited

Loading

virajjasani commented Mar 12, 2024 •

edited

Loading

virajjasani commented Mar 12, 2024 •

edited

Loading

virajjasani commented Mar 12, 2024 •

edited

Loading

virajjasani commented Mar 13, 2024

virajjasani commented Mar 13, 2024

steveloughran commented Mar 13, 2024

virajjasani commented Mar 14, 2024

steveloughran commented Mar 15, 2024

HADOOP-19066. S3A: AWS SDK V2 - Enabling FIPS should be allowed with central endpoint #6539

HADOOP-19066. S3A: AWS SDK V2 - Enabling FIPS should be allowed with central endpoint #6539

Conversation

virajjasani commented Feb 8, 2024

hadoop-yetus commented Feb 8, 2024

virajjasani commented Feb 9, 2024

virajjasani commented Mar 8, 2024

steveloughran left a comment

Choose a reason for hiding this comment

steveloughran commented Mar 12, 2024

steveloughran commented Mar 12, 2024

virajjasani commented Mar 12, 2024

steveloughran commented Mar 12, 2024 • edited Loading

virajjasani commented Mar 12, 2024 • edited Loading

virajjasani commented Mar 12, 2024 • edited Loading

virajjasani commented Mar 12, 2024 • edited Loading

virajjasani commented Mar 12, 2024 • edited Loading

virajjasani commented Mar 13, 2024

virajjasani commented Mar 13, 2024

steveloughran commented Mar 13, 2024

virajjasani commented Mar 14, 2024

steveloughran commented Mar 15, 2024

steveloughran commented Mar 12, 2024 •

edited

Loading

virajjasani commented Mar 12, 2024 •

edited

Loading

virajjasani commented Mar 12, 2024 •

edited

Loading

virajjasani commented Mar 12, 2024 •

edited

Loading

virajjasani commented Mar 12, 2024 •

edited

Loading