GoogleCloudStorageBlobStore.deleteBlobsIgnoringIfNotExists() does
not correctly catch StorageException thrown by batch.submit().
In the case a snapshot is deleted through BlobStoreRepository.deleteSnapshot()
a storage exception is not caught (only IOException are) so the deletion is
interrupted and indices cannot be cleaned up. The storage exception bubbles
up to SnapshotService.deleteSnapshotFromRepository() but the listener that
removes the deletion from the cluster state is not executed, leaving the
deletion in the cluster state.
This bug has been reported in #46772 where batch.submit() threw an
exception in the test testIndicesDeletedFromRepository and following
tests failed because a snapshot deletion was running.
Relates #46772
This commit adds support for Put Block API to the internal HTTP server
used in Azure repository integration tests. This allows to test the
behavior of the Azure SDK client when the Azure Storage service
returns errors when uploading Blob in multiple blocks or when
downloading a blob using ranged downloads.
This commit adds support for resumable uploads to the internal HTTP
server used in GoogleCloudStorageBlobStoreRepositoryTests. This
way we can also test the behavior of the Google's client when the
service returns server errors in response to resumable upload requests.
The BlobStore implementation for GCS has the choice between 2
methods to upload a blob: resumable and multipart. In the current
implementation, the client executes a resumable upload if the blob
size is larger than LARGE_BLOB_THRESHOLD_BYTE_SIZE,
otherwise it executes a multipart upload. This commit makes this
logic overridable in tests, allowing to randomize the decision of
using one method or the other.
The commit add support for single request resumable uploads
and chunked resumable uploads (the blob is uploaded into multiple
2Mb chunks; each chunk being a resumable upload). For this last
case, this PR also adds a test testSnapshotWithLargeSegmentFiles
which makes it more probable that a chunked resumable upload is
executed.
A resumable upload session can fail on with a 410 error and should
be retried in that case. I added retrying twice using resetting of
the given `InputStream` as the retry mechanism since the same
approach is used by the AWS S3 SDK already as well and relied upon
by the S3 repository implementation.
Related GCS documentation:
https://cloud.google.com/storage/docs/json_api/v1/status-codes#410_Gone
There were some issues with the Azure implementation requiring
permissions to list all containers ue to a container exists
check. This was caught in CI this time, but going forward we
should ensure that CI is executed using a token that does not
allow listing containers.
Relates #43288
Today if the connection to S3 times out or drops after starting to download an
object then the SDK does not attempt to recover or resume the download, causing
the restore of the whole shard to fail and retry. This commit allows
Elasticsearch to detect such a mid-stream failure and to resume the download
from where it failed.
This commit modifies the HTTP server used in
AzureBlobStoreRepositoryTests so that it randomly returns
server errors for any type of request executed by the Azure client.
This commit modifies the HTTP server used in
GoogleCloudStorageBlobStoreRepositoryTests so that it randomly
returns server errors. The test does not inject server errors for the
following types of request: batch request, resumable upload request.
The `repository-s3` plugin has supported a storage class of `onezone_ia` since
the SDK upgrade in #30723, but we do not test or document this fact. This
commit adds this storage class to the docs and adds a test to ensure that the
documented storage classes are all accepted by S3 too.
Fixes#30474
This commit removes the usage of MockGoogleCloudStoragePlugin in
GoogleCloudStorageBlobStoreRepositoryTests and replaces it by a
HttpServer that emulates the Storage service. This allows the repository
tests to use the real Google's client under the hood in tests and will allow
us to test the behavior of the snapshot/restore feature for GCS repositories
by simulating random server-side internal errors.
The HTTP server used to emulate the Storage service is intentionally simple
and minimal to keep things understandable and maintainable. Testing full
client options on the server side (like authentication, chunked encoding
etc) remains the responsibility of the GoogleCloudStorageFixture.
Similarly to what had been done for S3 (#46081) and GCS (#46255)
this commit adds repository integration tests for Azure, based on an
internal HTTP server instead of mocks.
When some high values are randomly picked up - for example the number
of indices to snapshot or the number of snapshots to create - the tests
in S3BlobStoreRepositoryTests can generate a high number of requests to
the internal S3 server.
In order to test the retry logic of the S3 client, the internal server is
designed to randomly generate random server errors. When many
requests are made, it is possible that the S3 client reaches its maximum
number of successive retries capacity. Then the S3 client will stop
retrying requests until enough retry attempts succeed, but it means
that any request could fail before reaching the max retries count and
make the test fail too.
Closes#46217Closes#46218Closes#46219
This commit modifies the HTTP server used in S3BlobStoreRepositoryTests
so that it randomly returns server errors for any type of request executed by
the SDK client. It is now possible to verify that the repository tests are s
uccessfully completed even if one or more errors were returned by the S3
service in response of a blob upload, a blob deletion or a object listing request
etc.
Because injecting errors forces the SDK client to retry requests, the test limits
the maximum errors to send in response for each request at 3 retries.
This commit removes the usage of MockAmazonS3 in S3BlobStoreRepositoryTests
and replaces it by a HttpServer that emulates the S3 service. This allows the
repository tests to use the real Amazon's S3 client under the hood in tests and will
allow to test the behavior of the snapshot/restore feature for S3 repositories by
simulating random server-side internal errors.
The HTTP server used to emulate the S3 service is intentionally simple and minimal
to keep things understandable and maintainable. Testing full client options on the
server side (like authentication, chunked encoding etc) remains the responsibility
of the AmazonS3Fixture.
This commit starts from the simple premise that the use of node settings
in blob store repositories is a mistake. Here we see that the node
settings are used to get default settings for store and restore throttle
rates. Yet, since there are not any node settings registered to this
effect, there can never be a default setting to fall back to there, and
so we always end up falling back to the default rate. Since this was the
only use of node settings in blob store repository, we move them. From
this, several places fall out where we were chaining settings through
only to get them to the blob store repository, so we clean these up as
well. That leaves us with the changeset in this commit.
This commit refactors the S3 credentials tests in
RepositoryCredentialsTests so that it now uses a single
node (ESSingleNodeTestCase) to test how secure/insecure
credentials are overriding each other. Using a single node
makes it much easier to understand what each test is actually
testing and IMO better reflect how things are initialized.
It also allows to fold into this class the test
testInsecureRepositoryCredentials which was wrongly located
in S3BlobStoreRepositoryTests. By moving this test away, the
S3BlobStoreRepositoryTests class does not need the
allow_insecure_settings option anymore and thus can be
executed as part of the usual gradle test task.
This commit changes the tests added in #45383 so that the fixture that
emulates the S3 service now sometimes consumes all the request body
before sending an error, sometimes consumes only a part of the request
body and sometimes consumes nothing. The idea here is to beef up a bit
the tests that writes blob because the client's retry logic relies on
marking and resetting the blob's input stream.
This pull request also changes the testWriteBlobWithRetries() so that it
(rarely) tests with a large blob (up to 1mb), which is more than the client's
default read limit on input streams (131Kb).
Finally, it optimizes the ZeroInputStream so that it is a bit more effective
(now works using an internal buffer and System.arraycopy() primitives).
This commit adds tests to verify the behavior of the S3BlobContainer and
its underlying AWS SDK client when the remote S3 service is responding
errors or not responding at all. The expected behavior is that requests are
retried multiple times before the client gives up and the S3BlobContainer
bubbles up an exception.
The test verifies the behavior of BlobContainer.writeBlob() and
BlobContainer.readBlob(). In the case of S3 writing a blob can be executed
as a single upload or using multipart requests; the test checks both scenario
by writing a small then a large blob.
* Repository Cleanup Endpoint (#43900)
* Snapshot cleanup functionality via transport/REST endpoint.
* Added all the infrastructure for this with the HLRC and node client
* Made use of it in tests and resolved relevant TODO
* Added new `Custom` CS element that tracks the cleanup logic.
Kept it similar to the delete and in progress classes and gave it
some (for now) redundant way of handling multiple cleanups but only allow one
* Use the exact same mechanism used by deletes to have the combination
of CS entry and increment in repository state ID provide some
concurrency safety (the initial approach of just an entry in the CS
was not enough, we must increment the repository state ID to be safe
against concurrent modifications, otherwise we run the risk of "cleaning up"
blobs that just got created without noticing)
* Isolated the logic to the transport action class as much as I could.
It's not ideal, but we don't need to keep any state and do the same
for other repository operations
(like getting the detailed snapshot shard status)
This change adds a new option called user_dictionary_rules to
Kuromoji's tokenizer. It can be used to set additional tokenization rules
to the Japanese tokenizer directly in the settings (instead of using a file).
This commit also adds a check that no rules are duplicated since this is not allowed
in the UserDictionary.
Closes#25343
* Update the REST API specification
This patch updates the REST API spefication in JSON files to better encode deprecated entities,
to improve specification of URL paths, and to open up the schema for future extensions.
Notably, it changes the `paths` from a list of strings to a list of objects, where each
particular object encodes all the information for this particular path: the `parts` and the `methods`.
Among the benefits of this approach is eg. encoding the difference between using the `PUT` and `POST`
methods in the Index API, to either use a specific document ID, or let Elasticsearch generate one.
Also `documentation` becomes an object that supports an `url` and also a `description` which is a
new field.
* Adapt YAML runner to new REST API specification format
The logic for choosing the path to use when running tests has been
simplified, as a consequence of the path parts being listed under each
path in the spec. The special case for create and index has been removed.
Also the parsing code has been hardened so that errors are thrown earlier
when the structure of the spec differs from what expected, and their
error messages should be more helpful.
The current implementations make it difficult for
adding new privileges (example: a cluster privilege which is
more than cluster action-based and not exposed to the security
administrator). On the high level, we would like our cluster privilege
either:
- a named cluster privilege
This corresponds to `cluster` field from the role descriptor
- or a configurable cluster privilege
This corresponds to the `global` field from the role-descriptor and
allows a security administrator to configure them.
Some of the responsibilities like the merging of action based cluster privileges
are now pushed at cluster permission level. How to implement the predicate
(using Automaton) is being now enforced by cluster permission.
`ClusterPermission` helps in enforcing the cluster level access either by
performing checks against cluster action and optionally against a request.
It is a collection of one or more permission checks where if any of the checks
allow access then the permission allows access to a cluster action.
Implementations of cluster privilege must be able to provide information
regarding the predicates to the cluster permission so that can be enforced.
This is enforced by making implementations of cluster privilege aware of
cluster permission builder and provide a way to specify how the permission is
to be built for a given privilege.
This commit renames `ConditionalClusterPrivilege` to `ConfigurableClusterPrivilege`.
`ConfigurableClusterPrivilege` is a renderable cluster privilege exposed
as a `global` field in role descriptor.
Other than this there is a requirement where we would want to know if a cluster
permission is implied by another cluster-permission (`has-privileges`).
This is helpful in addressing queries related to privileges for a user.
This is not just simply checking of cluster permissions since we do not
have access to runtime information (like request object).
This refactoring does not try to address those scenarios.
Relates #44048
Elasticsearch does not grant Netty reflection access to get Unsafe. The
only mechanism that currently exists to free direct buffers in a timely
manner is to use Unsafe. This leads to the occasional scenario, under
heavy network load, that direct byte buffers can slowly build up without
being freed.
This commit disables Netty direct buffer pooling and moves to a strategy
of using a single thread-local direct buffer for interfacing with sockets.
This will reduce the memory usage from networking. Elasticsearch
currently derives very little value from direct buffer usage (TLS,
compression, Lucene, Elasticsearch handling, etc all use heap bytes). So
this seems like the correct trade-off until that changes.
Uses JDK 11's per-socket configuration of TCP keepalive (supported on Linux and Mac), see
https://bugs.openjdk.java.net/browse/JDK-8194298, and exposes these as transport settings.
By default, these options are disabled for now (i.e. fall-back to OS behavior), but we would like
to explore whether we can enable them by default, in particular to force keepalive configurations
that are better tuned for running ES.
Currently in the transport-nio work we connect and bind channels on the
a thread before the channel is registered with a selector. Additionally,
it is at this point that we set all the socket options. This commit
moves these operations onto the event-loop after the channel has been
registered with a selector. It attempts to set the socket options for a
non-server channel at registration time. If that fails, it will attempt
to set the options after the channel is connected. This should fix
#41071.
* Stop Passing Around REST Request in Multiple Spots
* Motivated by #44564
* We are currently passing the REST request object around to a large number of places. This works fine since we simply copy the full request content before we handle the rest itself which is needlessly hard on GC and heap.
* This PR removes a number of spots where the request is passed around needlessly. There are many more spots to optimize in follow-ups to this, but this one would already enable bypassing the request copying for some error paths in a follow up.
* Create S3 Third Party Test Task that Covers the S3 CLI Tool
* Adjust snapshot cli test tool tests to work with real S3
* Build adjustment
* Clean up repo path before testing
* Dedup the logic for asserting path contents by using the correct utility method here that somehow became unused
We often start testing with early access versions of new Java
versions and this have caused minor issues in our tests
(i.e. #43141) because the version string that the JVM reports
cannot be parsed as it ends with the string -ea.
This commit changes how we parse and compare Java versions to
allow correct parsing and comparison of the output of java.version
system property that might include an additional alphanumeric
part after the version numbers
(see [JEP 223[(https://openjdk.java.net/jeps/223)). In short it
handles a version number part, like before, but additionally a
PRE part that matches ([a-zA-Z0-9]+).
It also changes a number of tests that would attempt to parse
java.specification.version in order to get the full version
of Java. java.specification.version only contains the major
version and is thus inappropriate when trying to compare against
a version that might contain a minor, patch or an early access
part. We know parse java.version that can be consistently
parsed.
Resolves#43141
These Azure tests have hard println statements which means we always see
these messages during configuration. Yet, there are unnecessary most of
the time. This commit changes them to use debug logging.
The repository-hdfs runners need to be disabled it in fips mode.
Testing done for all the tasks, dynamic created and static (integTest, integTestHa, integSecureTest, integSecureHaTest)
While the code works perfectly well for a single segment, it returns the wrong values for multiple segments. E.g. If we have 500 docs in one segment and if we want to get the doc id = 280 then data.advanceExact(topDocs.scoreDocs[i].doc) works fine. If we have two segments, say, with first segment having docs 1-200 and the second segment having docs 201-500, then 280 is fetched from the second segment but is actually 480. Subtracting the docBase (280-200) takes us to the correct document which is 80 in the second segment and actually 280.
Today we have an annotation for controlling logging levels in
tests. This annotation serves two purposes, one is to control the
logging level used in tests, when such control is needed to impact and
assert the behavior of loggers in tests. The other use is when a test is
failing and additional logging is needed. This commit separates these
two concerns into separate annotations.
The primary motivation for this is that we have a history of leaving
behind the annotation for the purpose of investigating test failures
long after the test failure is resolved. The accumulation of these stale
logging annotations has led to excessive disk consumption. Having
recently cleaned this up, we would like to avoid falling into this state
again. To do this, we are adding a link to the test failure under
investigation to the annotation when used for the purpose of
investigating test failures. We will add tooling to inspect these
annotations, in the same way that we have tooling on awaits fix
annotations. This will enable us to report on the use of these
annotations, and report when stale uses of the annotation exist.
Tanguy Leroux
Armin Braun
David Turner
Luca Cavanna
Mark Vieira
Henning Andersen
Jason Tedor
Jim Ferenczi
Igor Motov
Yogesh Gaikwad
Tim Brooks
Yannick Welsch
Ignacio Vera
Ioannis Kakavas
maarab7