Up to now a job update that reduces the model memory limit
was not allowed. However, there could definitely be cases
where reducing the limit is necessary and reasonable.
This commit makes it possible to decrease the limit as long
as it does not go below the current memory usage. We obtain
the latter from the model size stats.
The conditions under which updating the model_memory_limit
is not allowed are now:
- when the job is open
- latest model_size_stats.model_bytes < new value
relates elastic/x-pack-elasticsearch#2461
Original commit: elastic/x-pack-elasticsearch@5b35923590
Since elastic/x-pack-elasticsearch#3254 security headers have been stored in datafeed cluster state
to allow the datafeed to run searches using the credentials of the user
who created/updated it. As a result the parser was changed to read the
"headers" field so that cluster state could be reloaded. However, this
meant that datafeed configs could be submitted with a "headers" field.
No security loophole arose from this, as subsequent code overwrites the
contents of any supplied headers. But it could be confusing that an
erroneously supplied field did not cause a parse failure as it usually
would.
This change makes the config parser for datafeeds reject a "headers"
field. Now only the metadata parser used for reloading cluster state
will read a "headers" field.
Original commit: elastic/x-pack-elasticsearch@afa503275f
In order to more easily integrate xpack once it moves into the
elasticsearch repo, references to the existing x-pack-elasticsearch need
to be reduced. This commit introduces a few helper "methods" available
to any project within xpack (through gradle project extension
properties, as closures). All refeerences to project paths now use these
helper methods, except for those pertaining to bwc, which will be
handled in a followup.
Original commit: elastic/x-pack-elasticsearch@850668744c
This commit moves the source file in x-pack-core to a org.elasticsearch.xpack.core package. This is to prevent issues where we have compile-time success reaching through packages that will cross module boundaries at runtime (due to being in different classloaders). By moving these to a separate package, we have compile-time safety. Follow-ups can consider build time checking that only this package is defined in x-pack-core, or sealing x-pack-core until modules arrive for us.
Original commit: elastic/x-pack-elasticsearch@232e156e0e
Also, removes check for whether a job-to-remove exists
and replaces it with a check of whether a job-to-remove
is already present in the calendar. This allows to
remove a job that may no longer exists and it improves
feedback for the case that an existing job is removed from
a calendar that doesn't contain it.
relates elastic/x-pack-elasticsearch#3620
Original commit: elastic/x-pack-elasticsearch@3ea39be1b6
By moving tokenization for categorization to Java we give users access to considerably more options for tokenizing their log messages prior to using ML to categorize them. Now all Elasticsearch analyzer functionality is available, which opens up the possibility to sensibly categorize non-English log messages.
Relates elastic/machine-learning-cpp#491
Original commit: elastic/x-pack-elasticsearch@5d61b67614
* Use XPackRestIT as base class for XDocsClientYamlTestSuiteIT
* Remove the XPackRestTestCase class
* Address review comments
* Fix checkstyle checks
Original commit: elastic/x-pack-elasticsearch@c2a5e60c12
This change removes the InternalClient and the InternalSecurityClient. These are replaced with
usage of the ThreadContext and a transient value, `action.origin`, to indicate which component the
request came from. The security code has been updated to look for this value and ensure the
request is executed as the proper user. This work comes from elastic/x-pack-elasticsearch#2808 where @s1monw suggested
that we do this.
While working on this, I came across index template registries and rather than updating them to use
the new method, I replaced the ML one with the template upgrade framework so that we could
remove this template registry. The watcher template registry is still needed as the template must be
updated for rolling upgrades to work (see elastic/x-pack-elasticsearch#2950).
Original commit: elastic/x-pack-elasticsearch@7dbf2f263e
For the purpose of getting this API consumed by our UI, returning
overall buckets that match the job's largest `bucket_span` can
result in too much data. The UI only ever displays a few buckets
in the swimlane. Their span depends on the time range selected and
the screen resolution, but it will only ever be a relatively
low number.
This PR adds the ability to aggregate overall buckets in a user
specified `bucket_span`. That `bucket_span` may be equal or
greater to the largest job's `bucket_span`. The `overall_score`
of the result overall buckets is the max score of the
corresponding overall buckets with a span equal to the job's
largest `bucket_span`.
The implementation is now chunking the bucket requests
as otherwise the aggregation would fail when too many buckets
are matching.
Original commit: elastic/x-pack-elasticsearch@981f7a40e5
Adds the GET overall_buckets API.
The REST end point is: GET
/_xpack/ml/anomaly_detectors/job_id/results/overall_buckets
The API returns overall bucket results. An overall bucket
is a summarized bucket result over multiple jobs.
It has the `bucket_span` of the longest job's `bucket_span`.
It also has an `overall_score` that is the `top_n` average of the
max anomaly scores per job.
relates elastic/x-pack-elasticsearch#2693
Original commit: elastic/x-pack-elasticsearch@ba6061482d
The changes made for elastic/x-pack-elasticsearch#2369 showed that the ML security tests were seriously
weakened by the decision to grant many "minimal" privileges to all users
involved in the tests. A better solution is to override the auth header
such that a superuser runs setup actions and assertions that work by
querying raw documents in ways that an end user wouldn't. Then the ML
endpoints can be called with the privileges provided by the ML roles and
nothing else.
Original commit: elastic/x-pack-elasticsearch@4de42d9e54
Implementation details of ML endpoints should be performed using the
internal client, so that the end user only requires permissions for
the public ML endpoints and does not need to know how they are
implemented. This change fixes some instances where this rule was
not adhered to.
Original commit: elastic/x-pack-elasticsearch@01c8f5172c
* Add support for authz checks at on shard requests
* Add Rest Tests for authorization
* Bulk security - Only reject individual items, rather than a whole shard
* Sync with core change
* Grant "delete" priv in ML smoketest
This role had index and+bulk privileges but it also needs delete (in order to delete ML model-snapshots)
Original commit: elastic/x-pack-elasticsearch@830e89e652
This is related to elastic/x-pack-elasticsearch#1217. This PR removes the default password of
"changeme" from the reserved users.
This PR adds special behavior for authenticating the reserved users. No
ReservedRealm user can be authenticated until its password is set. The
one exception to this is the elastic user. The elastic user can be
authenticated with an empty password if the action is a rest request
originating from localhost. In this scenario where an elastic user is
authenticated with a default password, it will have metadata indicating
that it is in setup mode. An elastic user in setup mode is only
authorized to execute a change password request.
Original commit: elastic/x-pack-elasticsearch@e1e101a237
In does not make sense for the time_field in the data_description to
be used as a by/over/partition field name, nor the summary_count_field,
categorization_field or as an influencer. Therefore, configurations
where the time_field in the data_description is used in the
analysis_config are now rejected.
Additionally, it causes a problem communicating with the C++ code if
the control field name (which is '.') is used in the analysis_config,
so this is also rejected at the validation stage.
Relates elastic/x-pack-elasticsearch#1684
Original commit: elastic/x-pack-elasticsearch@e6750a2cda
In 5.4.x, the datafeed attempts to get all fields from
doc_values by default. It has a `_source` parameter which
when enabled changes the strategy to instead try to get
all fields from the source.
This has been the most common issue users have been
reporting as it means the datafeed will fail to fetch
any text fields by default.
This change uses the field capabilities API in order
to automatically detect whether a field is aggregatable.
It then extracts such fields from doc_values while the
rest are taken from source. The change also adds
validation to the start datafeed action so that if
fields are missing mappings or the time field is not
aggregatable we respond with an appropriate error.
relates elastic/x-pack-elasticsearch#1649
Original commit: elastic/x-pack-elasticsearch@76e2cc6cb2
* [ML] Not an error to close a job twice
* Error if job is opening
* Address review comments
* Test closed job isn’t resolved
Original commit: elastic/x-pack-elasticsearch@7da7b24c08
We didn't realise it was possible for a qa module to depend on the
test classes of the plugin module, so we duplicated a test class.
But it turns out it IS possible to declare this dependency and avoid
the duplication.
Original commit: elastic/x-pack-elasticsearch@b6a21cda28
This commit removes the SecuredString class that was previously used throughout the security code
and replaces it with the SecureString class from core that was added as part of the new secure
settings infrastructure.
relates elastic/x-pack-elasticsearch#421
Original commit: elastic/x-pack-elasticsearch@e9cd117ca1
Previously force closing a job required extra privileges. Following
the full discussion about what privileges should be required.
Original commit: elastic/x-pack-elasticsearch@4d85314b35
* Removed OPENING and CLOSING job states. Instead when persistent task has been created and
status hasn't been set then this means we haven't yet started, when the executor changes it to STARTED we have.
The coordinating node will monitor cs for a period of time until that happens and then returns or times out.
* Refactored job close api to go to node running job task and close job there.
* Changed unexpected job and datafeed exception messages to not mention the state and instead mention that job/datafeed haven't yet started/stopped.
Original commit: elastic/x-pack-elasticsearch@37e778b585
Dimitris Athanasiou
Tim Brooks
David Roberts
Ryan Ernst
Jason Tedor
Michael Basnight
David Kyle
Jay Modi
Tim Vernum
Martijn van Groningen
Hendrik Muhs