Introducing fs.azure.readahead.range parameter which can be set by the user.
Data will be populated in buffer for random reads as well which leads to fewer
remote calls.
This patch also changes the seek implementation to perform a lazy seek. The
actual seek is done when a read is initiated and data is not present in the buffer else
data is returned from the buffer thus reducing the number of remote storage calls.
Contributed By: Mukund Thakur
This addresses the regression in Hadoop 3.3.1 where if no S3 endpoint
is set in fs.s3a.endpoint, S3A filesystem creation may fail on
non-EC2 deployments, depending on the local host environment setup.
* If fs.s3a.endpoint is empty/null, and fs.s3a.endpoint.region
is null, the region is set to "us-east-1".
* If fs.s3a.endpoint.region is explicitly set to "" then the client
falls back to the SDK region resolution chain; this works on EC2
* Details in troubleshooting.md, including a workaround for Hadoop-3.3.1+
* Also contains some minor restructuring of troubleshooting.md
Contributed by Steve Loughran.
The S3A connector supports
"an auditor", a plugin which is invoked
at the start of every filesystem API call,
and whose issued "audit span" provides a context
for all REST operations against the S3 object store.
The standard auditor sets the HTTP Referrer header
on the requests with information about the API call,
such as process ID, operation name, path,
and even job ID.
If the S3 bucket is configured to log requests, this
information will be preserved there and so can be used
to analyze and troubleshoot storage IO.
Contributed by Steve Loughran.
The option `fs.s3a.endpoint.region` can be used
to explicitly set the AWS region of a bucket.
This is needed when using AWS Private Link, as
the region cannot be automatically determined.
Contributed by Mehakmeet Singh
When the S3A and ABFS filesystems are closed,
their IOStatistics are logged at debug in the log:
org.apache.hadoop.fs.statistics.IOStatisticsLogging
Set `fs.iostatistics.logging.level` to `info` for the statistics
to be logged at info. (also: `warn` or `error` for even higher
log levels).
Contributed by: Mehakmeet Singh
Followup to HADOOP-13327, which changed S3A output stream hsync/hflush calls
to raise an exception.
Adds a new option fs.s3a.downgrade.syncable.exceptions
When true, calls to Syncable hsync/hflush on S3A output streams will
log once at warn (for entire process life, not just the stream), then
increment IOStats with the relevant operation counter
With the downgrade option false (default)
* IOStats are incremented
* The UnsupportedOperationException current raised includes a link to the
JIRA.
Contributed by Steve Loughran.
The ABFS Filesystem and its input and output streams now implement
the IOStatisticSource interface and provide IOStatistics on
their interactions with Azure Storage.
This includes the min/max/mean durations of all REST API calls.
Contributed by Mehakmeet Singh <mehakmeet.singh@cloudera.com>
This moves the mock account name --which is required to never exist-- from
"mockAccount" to an account name containing a static UUID.
Contributed by Steve Loughran.
* HADOOP-16948. Support single writer dirs.
* HADOOP-16948. Fix findbugs and checkstyle problems.
* HADOOP-16948. Fix remaining checkstyle problems.
* HADOOP-16948. Add DurationInfo, retry policy for acquiring lease, and javadocs
* HADOOP-16948. Convert ABFS client to use an executor for lease ops
* HADOOP-16948. Fix ABFS lease test for non-HNS
* HADOOP-16948. Fix checkstyle and javadoc
* HADOOP-16948. Address review comments
* HADOOP-16948. Use daemon threads for ABFS lease ops
* HADOOP-16948. Make lease duration configurable
* HADOOP-16948. Add error messages to test assertions
* HADOOP-16948. Remove extra isSingleWriterKey call
* HADOOP-16948. Use only infinite lease duration due to cost of renewal ops
* HADOOP-16948. Remove acquire/renew/release lease methods
* HADOOP-16948. Rename single writer dirs to infinite lease dirs
* HADOOP-16948. Fix checkstyle
* HADOOP-16948. Wait for acquire lease future
* HADOOP-16948. Add unit test for acquire lease failure
Moves to the builder API for AWS S3 client creation, and
offers a similar style of API to the S3A FileSystem and tests, hiding
the details of which options are client, which are in AWS Conf,
and doing the wiring up of S3A statistics interfaces to the AWS
SDK internals. S3A Statistics, including IOStatistics, should now
count throttling events handled in the AWS SDK itself.
This patch restores endpoint determination by probes to US-East-1
if the client isn't configured with fs.s3a.endpoint.
Explicitly setting the endpoint will save the cost of these probe
HTTP requests.
Contributed by Steve Loughran.
The S3A connector's rename() operation now raises FileNotFoundException if
the source doesn't exist; a FileAlreadyExistsException if the destination
exists and is unsuitable for the source file/directory.
When renaming to a path which does not exist, the connector no longer checks
for the destination parent directory existing -instead it simply verifies
that there is no file immediately above the destination path.
This is needed to avoid race conditions with delete() and rename()
calls working on adjacent subdirectories.
Contributed by Steve Loughran.
Removed findbugs from the hadoop build images and added spotbugs instead.
Upgraded SpotBugs to 4.2.2 and spotbugs-maven-plugin to 4.2.0.
Reviewed-by: Masatake Iwasaki <iwasakims@apache.org>
Use spotbugs instead of findbugs. Removed findbugs from the hadoop build images,
and added spotbugs in the images instead.
Reviewed-by: Masatake Iwasaki <iwasakims@apache.org>
Reviewed-by: Inigo Goiri <inigoiri@apache.org>
Reviewed-by: Dinesh Chitlangia <dineshc@apache.org>
Adds an Abortable.abort() interface for streams to enable output streams to be terminated; this
is implemented by the S3A connector's output stream. It allows for commit protocols
to be implemented which commit/abort work by writing to the final destination and
using the abort() call to cancel any write which is not intended to be committed.
Consult the specification document for information about the interface and its use.
Contributed by Jungtaek Lim and Steve Loughran.
This defines what output streams and especially those which implement
Syncable are meant to do, and documents where implementations (HDFS; S3)
don't. With tests.
The file:// FileSystem now supports Syncable if an application calls
FileSystem.setWriteChecksum(false) before creating a file -checksumming
and Syncable.hsync() are incompatible.
Contributed by Steve Loughran.
The ABFS connector now implements listStatusIterator() with
asynchronous prefetching of the next page(s) of results.
For listing large directories this can provide tangible speedups.
If for any reason this needs to be disabled, set
fs.azure.enable.abfslistiterator to false.
Contributed by Bilahari T H.
When 403 is returned from an ABFS HTTP call, an AccessDeniedException is raised.
The exception text is unchanged, for any application string matching on the getMessage() contents.
Contributed by Steve Loughran.
* core-default.xml updated so that fs.s3a.committer.magic.enabled = true
* CommitConstants updated to match
* All tests which previously enabled the magic committer now rely on
default settings. This helps make sure it is enabled.
* Docs cover the switch, mention its enabled and explain why you may
want to disable it.
Note: this doesn't switch to using the committer -it just enables the path
rewriting magic which it depends on.
Contributed by Steve Loughran.
Caused by HADOOP-16830 and HADOOP-17271.
Fixes tests which fail intermittently based on configs and
in the case of the HugeFile tests, bulk runs with existing
FS instances meant statistic probes sometimes ended up probing those
of a previous FS.
Contributed by Steve Loughran.
Change-Id: I65ba3f44444e59d298df25ac5c8dc5a8781dfb7d
Caused by HADOOP-16380 and HADOOP-17271.
Fixes tests which fail intermittently based on configs and
in the case of the HugeFile tests, bulk runs with existing
FS instances meant statistic probes sometimes ended up probing those
of a previous FS.
Contributed by Steve Loughran.
Also fixes HADOOP-16995. ITestS3AConfiguration proxy tests failures when bucket probes == 0
The improvement should include the fix, ebcause the test would fail by default otherwise.
Change-Id: I9a7e4b5e6d4391ebba096c15e84461c038a2ec59
S3A connector to support the IOStatistics API of HADOOP-16830,
This is a major rework of the S3A Statistics collection to
* Embrace the IOStatistics APIs
* Move from direct references of S3AInstrumention statistics
collectors to interface/implementation classes in new packages.
* Ubiquitous support of IOStatistics, including:
S3AFileSystem, input and output streams, RemoteIterator instances
provided in list calls.
* Adoption of new statistic names from hadoop-common
Regarding statistic collection, as well as all existing
statistics, the connector now records min/max/mean durations
of HTTP GET and HEAD requests, and those of LIST operations.
Contributed by Steve Loughran.
DETAILS:
The previous commit for HADOOP-17397 was not the correct fix. DelegationSASGenerator.getDelegationSAS
should return sp=p for the set-permission and set-acl operations. The tests have also been updated as
follows:
1. When saoid and suoid are not specified, skoid must have an RBAC role assignment which grants
Microsoft.Storage/storageAccounts/blobServices/containers/blobs/modifyPermissions/action and sp=p
to set permissions or set ACL.
2. When saoid or suiod is specified, same as 1) but furthermore the saoid or suoid must be an owner of
the file or directory in order for the operation to succeed.
3. When saoid or suiod is specified, the ownership check is bypassed by also including 'o' (ownership)
in the SAS permission (for example, sp=op). Note that 'o' grants the saoid or suoid the ability to
change the file or directory owner to themself, and they can also change the owning group. Generally
speaking, if a trusted authorizer would like to give a user the ability to change the permissions or
ACL, then that user should be the file or directory owner.
TEST RESULTS:
namespace.enabled=true
auth.type=SharedKey
-------------------
$mvn -T 1C -Dparallel-tests=abfs -Dscale -DtestsThreadCount=8 clean verify
Tests run: 90, Failures: 0, Errors: 0, Skipped: 0
Tests run: 462, Failures: 0, Errors: 0, Skipped: 24
Tests run: 208, Failures: 0, Errors: 0, Skipped: 24
namespace.enabled=true
auth.type=OAuth
-------------------
$mvn -T 1C -Dparallel-tests=abfs -Dscale -DtestsThreadCount=8 clean verify
Tests run: 90, Failures: 0, Errors: 0, Skipped: 0
Tests run: 462, Failures: 0, Errors: 0, Skipped: 70
Tests run: 208, Failures: 0, Errors: 0, Skipped: 141
The addition of deprecated S3A configuration options in HADOOP-17318
triggered a reload of default (xml resource) configurations, which breaks
tests which fail if there's a per-bucket setting inconsistent with test
setup.
Creating an S3AFS instance before creating the Configuration() instance
for test runs gets that reload out the way before test setup takes
place.
Along with the fix, extra changes in the failing test suite to fail
fast when marker policy isn't as expected, and to log FS state better.
Rather than create and discard an instance, add a new static method
to S3AFS and invoke it in test setup. This forces the load
Change-Id: Id52b1c46912c6fedd2ae270e2b1eb2222a360329
This adds a semaphore to throttle the number of FileSystem instances which
can be created simultaneously, set in "fs.creation.parallel.count".
This is designed to reduce the impact of many threads in an application calling
FileSystem.get() on a filesystem which takes time to instantiate -for example
to an object where HTTPS connections are set up during initialization.
Many threads trying to do this may create spurious delays by conflicting
for access to synchronized blocks, when simply limiting the parallelism
diminishes the conflict, so speeds up all threads trying to access
the store.
The default value, 64, is larger than is likely to deliver any speedup -but
it does mean that there should be no adverse effects from the change.
If a service appears to be blocking on all threads initializing connections to
abfs, s3a or store, try a smaller (possibly significantly smaller) value.
Contributed by Steve Loughran.
This patch
* fixes the inversion
* adds a precondition check
* if the commands are supplied inverted, swaps them with a warning.
This is to stop breaking any tests written to cope with the existing
behavior.
Contributed by Steve Loughran
See also [SPARK-33402]: Jobs launched in same second have duplicate MapReduce JobIDs
Contributed by Steve Loughran.
Change-Id: Iae65333cddc84692997aae5d902ad8765b45772a
This fixes the S3Guard/Directory Marker Retention integration so that when
fs.s3a.directory.marker.retention=keep, failures during multipart delete
are handled correctly, as are incremental deletes during
directory tree operations.
In both cases, when a directory marker with children is deleted from
S3, the directory entry in S3Guard is not deleted, because it is still
critical to representing the structure of the store.
Contributed by Steve Loughran.
Change-Id: I4ca133a23ea582cd42ec35dbf2dc85b286297d2f
Unless you explicitly set it, the issue date of a delegation token identifier is 0, which confuses spark renewal (SPARK-33440). This patch makes sure that all S3A DT identifiers have the current time as issue date, fixing the problem as far as S3A tokens are concerned.
Contributed by Jungtaek Lim.
Fixes read-ahead buffer management issues introduced by HADOOP-16852,
"ABFS: Send error back to client for Read Ahead request failure".
Contributed by Sneha Vijayarajan
This reverts changes in HADOOP-13230 to use S3Guard TTL in choosing when
to issue a HEAD request; fixing tests to compensate.
New org.apache.hadoop.fs.s3a.performance.OperationCost cost,
S3GUARD_NONAUTH_FILE_STATUS_PROBE for use in cost tests.
Contributed by Steve Loughran.
Change-Id: I418d55d2d2562a48b2a14ec7dee369db49b4e29e
S3AFileSystem.listStatus() is optimized for invocations
where the path supplied is a non-empty directory.
The number of S3 requests is significantly reduced, saving
time, money, and reducing the risk of S3 throttling.
Contributed by Mukund Thakur.
Contributed by Sneha Vijayarajan
DETAILS:
This change adds config key "fs.azure.enable.conditional.create.overwrite" with
a default of true. When enabled, if create(path, overwrite: true) is invoked
and the file exists, the ABFS driver will first obtain its etag and then attempt
to overwrite the file on the condition that the etag matches. The purpose of this
is to mitigate the non-idempotency of this method. Specifically, in the event of
a network error or similar, the client will retry and this can result in the file
being created more than once which may result in data loss. In essense this is
like a poor man's file handle, and will be addressed more thoroughly in the future
when support for lease is added to ABFS.
TEST RESULTS:
namespace.enabled=true
auth.type=SharedKey
-------------------
$mvn -T 1C -Dparallel-tests=abfs -Dscale -DtestsThreadCount=8 clean verify
Tests run: 87, Failures: 0, Errors: 0, Skipped: 0
Tests run: 457, Failures: 0, Errors: 0, Skipped: 42
Tests run: 207, Failures: 0, Errors: 0, Skipped: 24
namespace.enabled=true
auth.type=OAuth
-------------------
$mvn -T 1C -Dparallel-tests=abfs -Dscale -DtestsThreadCount=8 clean verify
Tests run: 87, Failures: 0, Errors: 0, Skipped: 0
Tests run: 457, Failures: 0, Errors: 0, Skipped: 74
Tests run: 207, Failures: 0, Errors: 0, Skipped: 140
This changes directory tree deletion so that only files are incrementally deleted
from S3Guard after the objects are deleted; the directories are left alone
until metadataStore.deleteSubtree(path) is invoke.
This avoids directory tombstones being added above files/child directories,
which stop the treewalk and delete phase from working.
Also:
* Callback to delete objects splits files and dirs so that
any problems deleting the dirs doesn't trigger s3guard updates
* New statistic to measure #of objects deleted, alongside request count.
* Callback listFilesAndEmptyDirectories renamed listFilesAndDirectoryMarkers
to clarify behavior.
* Test enhancements to replicate the failure and verify the fix
Contributed by Steve Loughran
Adds the options to control the size of the per-output-stream threadpool
when writing data through the abfs connector
* fs.azure.write.max.concurrent.requests
* fs.azure.write.max.requests.to.queue
Contributed by Bilahari T H
Now skips ITestS3AEncryptionWithDefaultS3Settings.testEncryptionOverRename
when server side encryption is not set to sse:kms
Contributed by Mukund Thakur
This adds an option to disable "empty directory" marker deletion,
so avoid throttling and other scale problems.
This feature is *not* backwards compatible.
Consult the documentation and use with care.
Contributed by Steve Loughran.
Change-Id: I69a61e7584dc36e485d5e39ff25b1e3e559a1958
Contributed by Steve Loughran.
Fixes a condition which can cause job commit to fail if a task was
aborted < 60s before the job commit commenced: the task abort
will shut down the thread pool with a hard exit after 60s; the
job commit POST requests would be scheduled through the same pool,
so be interrupted and fail. At present the access is synchronized,
but presumably the executor shutdown code is calling wait() and releasing
locks.
Task abort is triggered from the AM when task attempts succeed but
there are still active speculative task attempts running. Thus it
only surfaces when speculation is enabled and the final tasks are
speculating, which, given they are the stragglers, is not unheard of.
Note: this problem has never been seen in production; it has surfaced
in the hadoop-aws tests on a heavily overloaded desktop
Contributed by Thomas Marquardt
DETAILS: WASB depends on the Azure Storage Java SDK. There is a concurrency
bug in the Azure Storage Java SDK that can cause the results of a list blobs
operation to appear empty. This causes the Filesystem listStatus and similar
APIs to return empty results. This has been seen in Spark work loads when jobs
use more than one executor core.
See Azure/azure-storage-java#546 for details on the bug in the Azure Storage SDK.
TESTS: A new test was added to validate the fix. All tests are passing:
wasb:
mvn -T 1C -Dparallel-tests=wasb -Dscale -DtestsThreadCount=8 clean verify
Tests run: 248, Failures: 0, Errors: 0, Skipped: 11
Tests run: 651, Failures: 0, Errors: 0, Skipped: 65
abfs:
mvn -T 1C -Dparallel-tests=abfs -Dscale -DtestsThreadCount=8 clean verify
Tests run: 64, Failures: 0, Errors: 0, Skipped: 0
Tests run: 437, Failures: 0, Errors: 0, Skipped: 33
Tests run: 206, Failures: 0, Errors: 0, Skipped: 24
Contributed by Thomas Marquardt.
DETAILS:
1) The authentication version in the service has been updated from Dec19 to Feb20, so need to update the client.
2) Add support and test cases for getXattr and setXAttr.
3) Update DelegationSASGenerator and related to use Duration instead of int for time periods.
4) Cleanup DelegationSASGenerator switch/case statement that maps operations to permissions.
5) Cleanup SASGenerator classes to use String.equals instead of ==.
TESTS:
Added tests for getXAttr and setXAttr.
All tests are passing against my account in eastus2euap:
$mvn -T 1C -Dparallel-tests=abfs -Dscale -DtestsThreadCount=8 clean verify
Tests run: 76, Failures: 0, Errors: 0, Skipped: 0
Tests run: 441, Failures: 0, Errors: 0, Skipped: 33
Tests run: 206, Failures: 0, Errors: 0, Skipped: 24
Contributed by Steve Loughran.
S3A delegation token providers will be asked for any additional
token issuers, an array can be returned,
each one will be asked for tokens when DelegationTokenIssuer collects
all the tokens for a filesystem.
Contributed by Mukund Thakur and Steve Loughran.
This patch ensures that writes to S3A fail when more than 10,000 blocks are
written. That upper bound still exists. To write massive files, make sure
that the value of fs.s3a.multipart.size is set to a size which is large
enough to upload the files in fewer than 10,000 blocks.
Change-Id: Icec604e2a357ffd38d7ae7bc3f887ff55f2d721a
Contributed by Thomas Marquardt.
DETAILS:
Previously we had a SASGenerator class which generated Service SAS, but we need to add DelegationSASGenerator.
I separated SASGenerator into a base class and two subclasses ServiceSASGenerator and DelegationSASGenreator. The
code in ServiceSASGenerator is copied from SASGenerator but the DelegationSASGenrator code is new. The
DelegationSASGenerator code demonstrates how to use Delegation SAS with minimal permissions, as would be used
by an authorization service such as Apache Ranger. Adding this to the tests helps us lock in this behavior.
Added a MockDelegationSASTokenProvider for testing User Delegation SAS.
Fixed the ITestAzureBlobFileSystemCheckAccess tests to assume oauth client ID so that they are ignored when that
is not configured.
To improve performance, AbfsInputStream/AbfsOutputStream re-use SAS tokens until the expiry is within 120 seconds.
After this a new SAS will be requested. The default period of 120 seconds can be changed using the configuration
setting "fs.azure.sas.token.renew.period.for.streams".
The SASTokenProvider operation names were updated to correspond better with the ADLS Gen2 REST API, since these
operations must be provided tokens with appropriate SAS parameters to succeed.
Support for the version 2.0 AAD authentication endpoint was added to AzureADAuthenticator.
The getFileStatus method was mistakenly calling the ADLS Gen2 Get Properties API which requires read permission
while the getFileStatus call only requires execute permission. ADLS Gen2 Get Status API is supposed to be used
for this purpose, so the underlying AbfsClient.getPathStatus API was updated with a includeProperties
parameter which is set to false for getFileStatus and true for getXAttr.
Added SASTokenProvider support for delete recursive.
Fixed bugs in AzureBlobFileSystem where public methods were not validating the Path by calling makeQualified. This is
necessary to avoid passing null paths and to convert relative paths into absolute paths.
Canonicalized the path used for root path internally so that root path can be used with SAS tokens, which requires
that the path in the URL and the path in the SAS token match. Internally the code was using
"//" instead of "/" for the root path, sometimes. Also related to this, the AzureBlobFileSystemStore.getRelativePath
API was updated so that we no longer remove and then add back a preceding forward / to paths.
To run ITestAzureBlobFileSystemDelegationSAS tests follow the instructions in testing_azure.md under the heading
"To run Delegation SAS test cases". You also need to set "fs.azure.enable.check.access" to true.
TEST RESULTS:
namespace.enabled=true
auth.type=SharedKey
-------------------
$mvn -T 1C -Dparallel-tests=abfs -Dscale -DtestsThreadCount=8 clean verify
Tests run: 63, Failures: 0, Errors: 0, Skipped: 0
Tests run: 432, Failures: 0, Errors: 0, Skipped: 41
Tests run: 206, Failures: 0, Errors: 0, Skipped: 24
namespace.enabled=false
auth.type=SharedKey
-------------------
$mvn -T 1C -Dparallel-tests=abfs -Dscale -DtestsThreadCount=8 clean verify
Tests run: 63, Failures: 0, Errors: 0, Skipped: 0
Tests run: 432, Failures: 0, Errors: 0, Skipped: 244
Tests run: 206, Failures: 0, Errors: 0, Skipped: 24
namespace.enabled=true
auth.type=SharedKey
sas.token.provider.type=MockDelegationSASTokenProvider
enable.check.access=true
-------------------
$mvn -T 1C -Dparallel-tests=abfs -Dscale -DtestsThreadCount=8 clean verify
Tests run: 63, Failures: 0, Errors: 0, Skipped: 0
Tests run: 432, Failures: 0, Errors: 0, Skipped: 33
Tests run: 206, Failures: 0, Errors: 0, Skipped: 24
namespace.enabled=true
auth.type=OAuth
-------------------
$mvn -T 1C -Dparallel-tests=abfs -Dscale -DtestsThreadCount=8 clean verify
Tests run: 63, Failures: 0, Errors: 0, Skipped: 0
Tests run: 432, Failures: 0, Errors: 1, Skipped: 74
Tests run: 206, Failures: 0, Errors: 0, Skipped: 140
Contributed by: Mehakmeet Singh
In some cases, ABFS-prefetch thread runs in the background which returns some bytes from the buffer and gives an extra readOp. Thus, making readOps values arbitrary and giving intermittent failures in some cases. Hence, readOps values of 2 or 3 are seen in different setups.
