Today either all nodes in the cluster connect to remote clusters of only nodes
that have remote clusters configured in their node config. To allow global remote
cluster configuration but restrict connections to a set of nodes in the cluster
this change adds a new setting `search.remote.connect` (defaults to `true`) to allow
to disable remote cluster connections on a per node basis.
In order to support the evolving GeoPoint encodings in Lucene 5 and 6, ES 2.x and 5.x implements an abstraction layer to the GeoPointFieldMapper classes. As of 5.x the geo_point field mapper settled on using Lucene's more performant LatLonPoint field type and deprecated all other encodings. In 6.0 all encodings except LatLonPoint have been removed rendering this abstraction layer useless. This commit removes the abstraction layer and renames the LatLonPointFieldMapper back to GeoPointFieldMapper to mantain consistency with ES field naming.
create a snapshot with a name that already exists in the repository.
Instead of throwing a SnapshotCreateException, which results in a
generic 500 status code, a duplicate snapshot name will throw a
InvalidSnapshotNameException, which will result in a 400 status code
(bad request).
`QUERY_AND_FETCH` has been treated as an internal optimization for 2 major
versions. This commit removes the search type and it's implementation details and
folds the optimization in the case of a single shard into the search controller such
that every search with a single shard (non DFS) will receive this optimization.
When a node receives a new cluster state from the master, it opens up connections to any new node in the cluster state. That has always been done serially on the cluster state thread but it has been a long standing TODO to do this concurrently, which is done by this PR.
This is spin off of #22828, where an extra handshake is done whenever connecting to a node, which may slow down connecting. Also, the handshake is done in a blocking fashion which triggers assertions w.r.t blocking requests on the cluster state thread. Instead of adding an exception, I opted to implement concurrent connections which both side steps the assertion and compensates for the extra handshake.
This changes the way that replica failures are handled such that not all
failures will cause the replica shard to be failed or marked as stale.
In some cases such as refresh operations, or global checkpoint syncs, it is
"okay" for the operation to fail without the shard being failed (because no data
is out of sync). In these cases, instead of failing the shard we should simply
fail the operation, and, in the event it is a user-facing operation, return a
5xx response code including the shard-specific failures.
This was accomplished by having two forms of the `Replicas` proxy, one that is
for non-write operations that does not fail the shard, and one that is for write
operations that will fail the shard when an operation fails.
Relates to #10708
It was accidentally renamed `enabled_position_increment` in the cleanups
for 5.0. This adds `enable_position_increment` as a deprecated alias
so it will continue to work.
This commit removes the following queries and parameters (which were deprecated in 5.0):
* GeoPointDistanceRangeQuery
* coerce, and ignore_malformed for GeoBoundingBoxQuery, GeoDistanceQuery, GeoPolygonQuery, and GeoDistanceSort
This is related to #22116. Core no longer needs `SocketPermission`
`connect`.
This permission is relegated to these modules/plugins:
- transport-netty4 module
- reindex module
- repository-url module
- discovery-azure-classic plugin
- discovery-ec2 plugin
- discovery-gce plugin
- repository-azure plugin
- repository-gcs plugin
- repository-hdfs plugin
- repository-s3 plugin
And for tests:
- mocksocket jar
- rest client
- httpcore-nio jar
- httpasyncclient jar
This commit upgrades the checkstyle configuration from version 5.9 to
version 7.5, the latest version as of today. The main enhancement
obtained via this upgrade is better detection of redundant modifiers.
Relates #22960
When a primary is relocated from an old node to a new node, it can have
ops in its translog that do not have a sequence number assigned. When a
file-based recovery is started, this can lead to skipping these ops when
replaying the translog due to a bug in the recovery logic. This commit
addresses this bug and adds a test in the BWC tests.
Relates #22945
This change also removes the reference to the difference bewteen full name and index name.
They are always the same since 2.x and `name` does not refer anymore to `author.name` automatically.
A simple pattern must be used instead.
Remove redundant code that checks the field name twice.
Today if a user invokes the remove plugin command without specifying the
name of a plugin to remove, we arrive at a null pointer exception. This
commit adds logic to cleanly handle this situation and provide clear
feedback to the user.
Relates #22930
Currently, update action internally uses deprecated index and delete
transport actions. As of #21964, these tranport actions were deprecated
in favour of using single item bulk request. In this commit, update action
uses single item bulk action.
This change adds a strict mode for xcontent parsing on the rest layer. The strict mode will be off by default for 5.x and in a separate commit will be enabled by default for 6.0. The strict mode, which can be enabled by setting `http.content_type.required: true` in 5.x, will require that all incoming rest requests have a valid and supported content type header before the request is dispatched. In the non-strict mode, the Content-Type header will be inspected and if it is not present or not valid, we will continue with auto detection of content like we have done previously.
The content type header is parsed to the matching XContentType value with the only exception being for plain text requests. This value is then passed on with the content bytes so that we can reduce the number of places where we need to auto-detect the content type.
As part of this, many transport requests and builders were updated to provide methods that
accepted the XContentType along with the bytes and the methods that would rely on auto-detection have been deprecated.
In the non-strict mode, deprecation warnings are issued whenever a request with body doesn't provide the Content-Type header.
See #19388
GeoDistance query, sort, and scripts make use of a crazy GeoDistance enum for handling 4 different ways of computing geo distance: SLOPPY_ARC, ARC, FACTOR, and PLANE. Only two of these are necessary: ARC, PLANE. This commit removes SLOPPY_ARC, and FACTOR and cleans up the way Geo distance is computed.
This commit change ElasticsearchException.failureFromXContent() method so that it now parses root causes which were ignored before, and adds them as suppressed exceptions of the returned exception.
Implemented by wrapping an array of reused `ModuleDateTime`s that
we grow when needed. The `ModuleDateTime`s are reused when we
move to the next document.
Also improves the error message returned when attempting to modify
the `ScriptdocValues`, removes a couple of allocations, and documents
that the date functions are available in Painless.
Relates to #22162
This disallows object mappings that would accidentally create something like
`foo..bar`, which is then unparsable for the `bar` field as it does not know
what its parent is.
Resolves#22794
The test tried to create a situation where a stale replica is the only shard available. It did so by stopping the node with the replica, indexing some, stopping the primary node, starting a new node. This is flawed because the newly started node may reuse the data path of the primary node and things go back to green. Instead we should make sure that the replica is on the path that will be selected when the new node is started (i.e., the path with the smaller ordinal)
This commit adds a BytesRestResponse.errorFromXContent() method to parse the error returned by BytesRestResponse. It returns a ElasticsearchStatusException instance.
Currently, if a previously allocated shard has no in-sync copy in the
cluster, but there is a stale replica copy, the explain API does not
include information about the stale replica copies in its output. This
commit includes any shard copy information available (even for stale
copies) when explaining an unassigned primary shard that was previously
allocated in the cluster.
This situation can arise as follows: imagine an index with 1 primary and
1 replica and a cluster with 2 nodes. If the node holding the replica
is shut down, and data continues to be indexed, only the primary will
have the latest data and the replica that has gone offline will be
marked as stale. Now, suppose the node holding the primary is shut
down. There are no copies of the shard data in the cluster. Now, start
the first stopped node (holding the stale replica) back up. The cluster
is red because there is no in-sync copy available. Running the explain
API before would inform the user that there is no valid shard copy in
the cluster for that shard, but it would not provide any information
about the existence of the stale replica that exists on the restarted
node. With this commit, the explain API provides information about all
the stale replica copies when explaining the unassigned primary.
Currently, stored scripts use a namespace of (lang, id) to be put, get, deleted, and executed. This is not necessary since the lang is stored with the stored script. A user should only have to specify an id to use a stored script. This change makes that possible while keeping backwards compatibility with the previous namespace of (lang, id). Anywhere the previous namespace is used will log deprecation warnings.
The new behavior is the following:
When a user specifies a stored script, that script will be stored under both the new namespace and old namespace.
Take for example script 'A' with lang 'L0' and data 'D0'. If we add script 'A' to the empty set, the scripts map will be ["A" -- D0, "A#L0" -- D0]. If a script 'A' with lang 'L1' and data 'D1' is then added, the scripts map will be ["A" -- D1, "A#L1" -- D1, "A#L0" -- D0].
When a user deletes a stored script, that script will be deleted from both the new namespace (if it exists) and the old namespace.
Take for example a scripts map with {"A" -- D1, "A#L1" -- D1, "A#L0" -- D0}. If a script is removed specified by an id 'A' and lang null then the scripts map will be {"A#L0" -- D0}. To remove the final script, the deprecated namespace must be used, so an id 'A' and lang 'L0' would need to be specified.
When a user gets/executes a stored script, if the new namespace is used then the script will be retrieved/executed using only 'id', and if the old namespace is used then the script will be retrieved/executed using 'id' and 'lang'
The seq# base recovery logic relies on rolling back lucene to remove any operations above the global checkpoint. This part of the plan is not implemented yet but have to have these guarantees. Instead we should make the seq# logic validate that the last commit point (and the only one we have) maintains the invariant and if not, fall back to file based recovery.
This commit adds a test that creates situation where rollback is needed (primary failover with ops in flight) and fixes another issue that was surfaced by it - if a primary can't serve a seq# based recovery request and does a file copy, it still used the incoming `startSeqNo` as a filter.
Relates to #22484 & #10708
With the new secure settings, methods like getAsMap() no longer work
correctly as a means of checking for empty settings, or the total size.
This change converts the existing uses of that method to use methods
directly on Settings. Note this does not update the implementations to
account for SecureSettings, as that will require a followup which
changes how secure settings work.
* Integrate UnifiedHighlighter
This change integrates the Lucene highlighter called "unified" in the list of supported highlighters for ES.
This highlighter can extract offsets from either postings, term vectors, or via re-analyzing text.
The best strategy is picked automatically at query time and depends on the field and the query to highlight.
In #22762, settings preparation during bootstrap was changed slightly to
account for SecureSettings, by starting with a fresh settings builder
after reading the initial configuration. However, this the defaults from
system properties were never re-read. This change fixes that bug (which
was never released).
closes#22861
Usually the order in which we serialize sets and maps of things doesn't matter,
but since InnerHitBuilder is part of SearchSourceBuilder, which is in turn used
as a cache key in its bytes serialization, we need to ensure the order of all
these fields when writing them to an output stream.
This adds tests and makes sure we iterate over the scriptField set and the
childInnerHits map in a fixed order.
Closes#22808
Also adds many `equals` and `hashCode` implementations and moves
the failure printing in `MatchAssertion` into a common spot and
exposes it over `assertEqualsWithErrorMessageFromXContent` which
does an object equality test but then uses `toXContent` to print
the differences.
Relates to #22278
This moves the building blocks for delete by query into core. This
should enabled two thigns:
1. Plugins other than reindex to implement "bulk by scroll" style
operations.
2. Plugins to directly call delete by query. Those plugins should
be careful to make sure that task cancellation still works, but
this should be possible.
Notes:
1. I've mostly just moved classes and moved around tests methods.
2. I haven't been super careful about cohesion between these core
classes and reindex. They are quite interconnected because I wanted
to make the change as mechanical as possible.
Closes#22616
* S3 repository: Add named configurations
This change implements named configurations for s3 repository as
proposed in #22520. The access/secret key secure settings which were
added in #22479 are reverted, and the only secure settings are those
with the new named configs. All other previously used settings for the
connection are deprecated.
closes#22520
Also adds many `equals` and `hashCode` implementations and moves
the failure printing in `MatchAssertion` into a common spot and
exposes it over `assertEqualsWithErrorMessageFromXContent` which
does an object equality test but then uses `toXContent` to print
the differences.
Relates to #22278
This commit introduces sequence-number-based recovery. When a replica
has fallen out of sync, rather than performing a file-based recovery we
first attempt to replay operations since the last local checkpoint on
the replica. To do this, at the start of recovery the replica tells the
primary what its local checkpoint is. The primary will then wait for all
operations between that local checkpoint and the current maximum
sequence number to complete; this is to ensure that there are no gaps in
the operations that will be replayed from the primary to the
replica. This is a best-effort attempt as we currently have no
guarantees on the primary that these operations will be available; if we
are not able to replay all operations in the desired range, we just
fallback to file-based recovery. Later work will strengthen the
guarantees.
Relates #22484
At this point AbstractSearchAsyncAction is just a base-class for the first phase of a search where we have multiple replicas
for each shardID. If one of them is not available we move to the next one. Yet, once we passed that first stage we have to work with
the shards we succeeded on the initial phase.
Unfortunately, subsequent phases are not fully detached from the initial phase since they are all non-static inner classes.
In future changes this will be changed to detach the inner classes to test them in isolation and to simplify their creation.
The AbstractSearchAsyncAction should be final and it should just get a factory for the next phase instead of requiring subclasses
etc.
This commit adds a ElasticsearchException.failureFromXContent() that can be used to parse the result of ElasticsearchException.generateFailureXContent().
Today we cache query results even if the query timed out. This is obviously
problematic since results are not complete. Yet, the decision if a query timed
out or not happens too late to simply not cache the result since if we'd just throw
an exception all currently waiting requests with the same request / cache key would
fail with the same exception without the option to access the result or to re-execute.
Instead, this change will allow the request to enter the cache but invalidates it immediately.
Concurrent request might not get executed and return the timed out result which is not absolutely
correct but very likely since identical requests will likely timeout as well. As a side-effect
we won't hammer the node with concurrent slow searches but rather only execute one of them
and return shortly cached result.
Closes#22789
The output of the ElasticsearchException.generateThrowableXContent() method can be parsed back by the ElasticsearchException.fromXContent() method.
This commit adds unit tests in the style of the to-and-from-xcontent tests we already have for other parsing methods. It also relax the strict parsing of the ElasticsearchException.fromXContent() so that it does not throw an exception when custom metadata and headers are parsed, as long as they are either strings or arrays of strings. Every other type is ignored at parsing time.
Some tests verify that all connection have been closed but due to the
async / concurrent nature of `RemoteClusterConnection` there are situations
where we notify listeners that trigger tests to finish before we actually
closed all connections. The race is very very small and has no impact on the
code correctness. This commit documents and improves the way we close
connections to ensure test won't fail with false positives.
Closes#22803
This commit removes the search type `dfs_query_and_fetch` without a
replacement. We don't allow to use this type via REST since 2.x
but still keep it around for no particular reason. There we no users
complaining about the availability. This should now be removed from the
codebase. `query_and_fetch` is still used internally to safe a roundtrip
if there is only one shard but it can't be used via the rest interface.
This is related to #22116. URLRepository requires SocketPermission
connect. This commit introduces a new module called "repository-url"
where URLRepository will reside. With the new module, permissions can
be removed from core.
Add unit tests for `TopHitsAggregator` and convert some snippets in
docs for `top_hits` aggregation to `// CONSOLE`.
Relates to #22278
Relates to #18160
Previous to his change when the range query was rewritten to an unbounded range (`[* TO *]`) it maintained the timezone and format for the query. This means that queries with different timezones and format which are rewritten to unbounded range queries actually end up as different entries in the search request cache.
This is inefficient and unnecessary so this change nulls the timezone and format in the rewritten query so that regardless of the timezone or format the rewritten query will be the same.
Although this does not fix#22412 (since it deals with the WITHIN case rather than the INTERSECTS case) it is born from the same arguments
5.3.0 is the first version that contains the split from headers to metadata, updated the check to reflect that. It was previously after to be able to commit to master first, and only after that backport the change. Otherwise master tests would have failed until the change was backported.
Previously the behavior where the `OpType` byte was serialized was only in
master, but it was recently backported to 5.x, so the serialization version
checks need to be updated as well.
move "es." internal headers to separate metadata set in ElasticsearchException and stop returning them as response headers
Closes#17593
* [TEST] remove ESExceptionTests, move its methods to ElasticsearchExceptionTests or ExceptionSerializationTests
Docker cgroups are mounted in the wrong place (i.e., inconsistently with
/proc/self/cgroup). This commit adds an undocumented hack for working
around, for now.
Relates #22757
As a follow up to #22649, this changes the resent tests for parsing parts of search
responses to randomly set the humanReadable() flag of the XContentBuilder that
is used to render the responses. This should help to test that we can parse back
thoses classes if the user specifies `?human=true` in the request url.
There was a typo in the `ParseField` declaration. I know
we want to port these parsers to `ObjectParser` eventually
but I don't have the energy for that today and want to get
this fixed.
Closes#22722
* Add top hits collapsing to search request
The field collapsing is done with a custom top docs collector that "collapse" search hits with same field value.
The distributed aspect is resolve using the two passes that the regular search uses. The first pass "collapse" the top hits, then the coordinating node merge/collapse the top hits from each shard.
```
GET _search
{
"collapse": {
"field": "category",
}
}
```
This change also adds an ExpandCollapseSearchResponseListener that intercepts the search response and expands collapsed hits using the CollapseBuilder#innerHit} options.
The retrieval of each inner_hits is done by sending a query to all shards filtered by the collapse key.
```
GET _search
{
"collapse": {
"field": "category",
"inner_hits": {
"size": 2
}
}
}
```
To effectively allow a plugin to intercept a transport handler it needs
to know if the handler must be executed even if there is a rejection on the
thread pool in the case the wrapper forks a thread to execute the actual handler.
This commit adds a SpecialPermission constant and uses that constant
opposed to introducing new instances everywhere.
Additionally, this commit introduces a single static method to check that
the current code has permission. This avoids all the duplicated access
blocks that exist currently.
Today `InternalEngine#innerIndex` is a pretty big method (> 150 SLoC). This
commit merged `#index` and `#innerIndex` and splits it up into smaller contained
methods.
* Upgrade to Lucene 6.4.0
`ValueSource`s are now converted to `DoubleValueSource`s using the Lucene adapter made for the migration to the new API in 6.4.0.
Introduce CheckedSupplier and CheckedRunnable functional interfaces
into core. These offer a checked version of the Supplier and Runnable
interfaces for use with lambda apis.
This commit ensures that the index.latest blob is first examined to
determine the latest index-N blob id, before attempting to list all
index-N blobs and picking the blob with the highest N.
It also fixes the MockRepository#move so that tests are able to handle
non-atomic moves. This is done by adding a special setting to the
MockRepository that requires the test to specify if it can handle
non-atomic moves. If so, then the MockRepository#move operation will be
non-atomic to allow testing for against such repositories.
The weight factor function does not check if the delegate score function needs to access the score of the query.
This results in a _score equals to 0 for all score function that set a weight.
This change modifies the WeightFactorFunction#needsScore to delegate the call to its underlying score function.
Fix#21483
There are presently 7 ctor args used in any rest handlers:
* `Settings`: Every handler uses it to initialize a logger and
some other strange things.
* `RestController`: Every handler registers itself with it.
* `ClusterSettings`: Used by `RestClusterGetSettingsAction` to
render the default values for cluster settings.
* `IndexScopedSettings`: Used by `RestGetSettingsAction` to get
the default values for index settings.
* `SettingsFilter`: Used by a few handlers to filter returned
settings so we don't expose stuff like passwords.
* `IndexNameExpressionResolver`: Used by `_cat/indices` to
filter the list of indices.
* `Supplier<DiscoveryNodes>`: Used to fill enrich the response
by handlers that list tasks.
We probably want to reduce these arguments over time but
switching construction away from guice gives us tighter
control over the list of available arguments.
These parameters are passed to plugins using
`ActionPlugin#initRestHandlers` which is expected to build and
return that handlers immediately. This felt simpler than
returning an reference to the ctors given all the different
possible args.
Breaks java plugins by moving rest handlers off of guice.
Today we try to be smart and make a generic decision if an exception should
be treated as a document failure but in some cases concurrency in the index writer
make this decision very difficult since we don't have a consistent state in the case
another thread is currently failing the IndexWriter/InternalEngine due to a tragic event.
This change simplifies the exception handling and makes specific decisions about document failures
rather than using a generic heuristic. This prevent exceptions to be treated as document failures
that should have failed the engine but backed out of failing since since some other thread has
already taken over the failure procedure but didn't finish yet.
Similar to the Filters aggregation but only supports "keyed" filter buckets and automatically "ANDs" pairs of filters to produce a form of adjacency matrix.
The intersection of buckets "A" and "B" is named "A&B" (the choice of separator is configurable). Empty intersection buckets are removed from the final results.
Closes#22169
This is related to #22116. Core no longer needs SocketPermission
accept. This permission is relegated to the transport-netty4 module
and (for tests) to the mocksocket jar.
Those services validate their setting before submitting an AckedClusterStateUpdateTask to the cluster state service. An acked cluster state may be completed by a networking thread when the last acks as received. As such it needs special care to make sure that thread context headers are handled correctly.
This commit fixes an issue with deprecation logging for lenient
booleans. The underlying issue is that adding deprecation logging for
lenient booleans added a static deprecation logger to the Settings
class. However, the Settings class is initialized very early and in CLI
tools can be initialized before logging is initialized. This leads to
status logger error messages. Additionally, the deprecation logging for
a lot of the settings does not provide useful context (for example, in
the token filter factories, the deprecation logging only produces the
name of the setting, but gives no context which token filter factory it
comes from). This commit addresses both of these issues by changing the
call sites to push a deprecation logger through to the lenient boolean
parsing.
Relates #22696
In preparation for being able to parse SearchResponse from its rest representation
for the java rest client, this adds fromXContent to SearchProfileShardResults and its
nested classes.
Adds unit tests for the `filters` aggregation.
This change also adds an helper to search and reduce any aggregator in a unit test.
This is done by dividing a single searcher in sub-searcher, one for each segment.
Relates #22278
* Fix NPE on FieldStats with mixed cluster on version pre/post 5.2
In 5.2 the FieldStats API can return null min/max values.
These values cannot be deserialized by a node with version pre 5.2 so if this node
is pick to coordinate a FieldStats request in a mixed cluster an NPE can be thrown.
This change prevents the NPE by removing the non serializable FieldStats object directly in the field stats shard request.
The filtered fields will not be present in the response when a node pre 5.2 acts as a coordinating node.
This change is a simple adaptation of https://github.com/elastic/elasticsearch/pull/19587 for the current state of master.
It allows to define search response listener in the form of `BiConsumer<SearchRequest, SearchResponse>`s in a search plugin.
This PR removes all leniency in the conversion of Strings to booleans: "true"
is converted to the boolean value `true`, "false" is converted to the boolean
value `false`. Everything else raises an error.
Changes the error message when `action.auto_create_index` or
`index.mapper.dynamic` forbids automatic creation of an index
from `no such index` to one of:
* `no such index and [action.auto_create_index] is [false]`
* `no such index and [index.mapper.dynamic] is [false]`
* `no such index and [action.auto_create_index] contains [-<pattern>] which forbids automatic creation of the index`
* `no such index and [action.auto_create_index] ([all patterns]) doesn't match`
This should make it more clear *why* there is `no such index`.
Closes#22435
testAckedIndexing now waits for all nodes to stabilize in the cluster
state through an assertBusy before final validation that all documents
are found in tehir respective shards in the cluster. Before, what could
happen is that the ensureGreen check passes but only after that is a
ping failure from the network disruption processed by the master,
thereby rendering the cluster RED again. This assertBusy waits up to 30
seconds for all nodes to have stabilized and all get document actions to
succeed.
Today we do not preserve response headers if they are present on a transport protocol
response. While preserving these headers is not always desired, in the most cases we
should pass on these headers to have consistent results for depreciation headers etc.
yet, this hasn't been much of a problem since most of the deprecations are detected early
ie. on the coordinating node such that this bug wasn't uncovered until #22647
This commit allow to optionally preserve headers when a context is restored and also streamlines
the context restore since it leaked frequently into the callers thread context when the callers
context wasn't restored again.
Relates to #22024
On top of documentation, the PR adds deprecation loggers and deals with the resulting warning headers.
The yaml test is set exclude versions up to 6.0. This is need to make sure bwc tests pass until this is backported to 5.2.0 . Once that's done, I will change the yaml test version limits
This change makes it possible for custom routing values to go to a subset of shards rather than
just a single shard. This enables the ability to utilize the spatial locality that custom routing can
provide while mitigating the likelihood of ending up with an imbalanced cluster or suffering
from a hot shard.
This is ideal for large multi-tenant indices with custom routing that suffer from one or both of
the following:
- The big tenants cannot fit into a single shard or there is so many of them that they will likely
end up on the same shard
- Tenants often have a surge in write traffic and a single shard cannot process it fast enough
Beyond that, this should also be useful for use cases where most queries are done under the context
of a specific field (e.g. a category) since it gives a hint at how the data can be stored to minimize
the number of shards to check per query. While a similar solution can be achieved with multiple
concrete indices or aliases per value today, those approaches breakdown for high cardinality fields.
A partitioned index enforces that mappings have routing required, that the partition size does not
change when shrinking an index (the partitions will shrink proportionally), and rejects mappings
that have parent/child relationships.
Closes#21585
Instead of forcing each task to register all nodes where its children are running, this commit runs cancellation on all nodes. The task cancellation operation doesn't run too frequently, so this optimization doesn't seem to be worth additional complexity of the interface.
Previously, certain settings that could take multiple comma delimited
values would pick up incorrect values for all entries but the first if
each comma separated value was followed by a whitespace character. For
example, the multi-value "A,B,C" would be correctly parsed as
["A", "B", "C"] but the multi-value "A, B, C" would be incorrectly parsed
as ["A", " B", " C"].
This commit allows a comma separated list to have whitespace characters
after each entry. The specific settings that were affected by this are:
cluster.routing.allocation.awareness.attributes
index.routing.allocation.require.*
index.routing.allocation.include.*
index.routing.allocation.exclude.*
cluster.routing.allocation.require.*
cluster.routing.allocation.include.*
cluster.routing.allocation.exclude.*
http.cors.allow-methods
http.cors.allow-headers
For the allocation filtering related settings, this commit also provides
validation of each specified entry if the filtering is done by _ip,
_host_ip, or _publish_ip, to ensure that each entry is a valid IP
address.
Closes#22297
This commit tries to simplify the way ElasticsearchException are rendered to xcontent. It adds some documentation and renames and merges some methods. Current behavior is preserved, the goal is to be more readable and centralize everything in the ElasticsearchException class.
`EngineClosedException` is a ES level exception that is used to indicate that the engine is closed when operation starts. It doesn't really add much value and we can use `AlreadyClosedException` from Lucene (which may already bubble if things go wrong during operations). Having two exception can just add confusion and lead to bugs, like wrong handling of `EngineClosedException` when dealing with document level failures. The latter was exposed by `IndexWithShadowReplicasIT`.
This PR also removes the AwaitFix from the `IndexWithShadowReplicasIT` tests (which was what cause this to be discovered). While debugging the source of the issue I found some mismatches in document uid management in the tests. The term that was passed to the engine didn't correspond to the uid in the parsed doc - those are fixed as well.
Today we have quite some abstractions that are essentially providing a simple
dispatch method to the plugins defining a `HttpServerTransport`. This commit
removes `HttpServer` and `HttpServerAdaptor` and introduces a simple `Dispatcher` functional
interface that delegate to `RestController` by default.
Relates to #18482
#22025 deprecated this setting (pending it's removal) but it's frequent usage will spam the deprecation logs and also fails test. As temporary work around we should not use the setting object directly.
Currently both ProfileResult and CollectorResult print the time field in a human readable string format
(e.g. "time": "55.20315000ms"). When trying to parse this back to a long value, for example to use in
the planned high level java rest client, we can lose precision because of conversion and rounding issues.
This change adds a new additional field (`time_in_nanos`) to the profile response to be able to get the
original time value in nanoseconds back.
The old `time` field is only printed when the `?`human=true` flag in the url is set. This follow the behaviour for
all other stats-related apis. Also the format of the `time` field is slightly changed. Instead of always formatting
the output as a 10-digit ms value, by using the `XContentBuilder#timeValueField()` method we now print
the largest time unit present is used (e.g. "s", "ms", "micros").
An operation that completed successfully on a primary can result in a
version conflict on a replica due to the asynchronous nature of
operations. When a replica operation results in a version conflict, the
operation is not added to the translog. This leads to gaps in the
translog which is problematic as it can lead to situations where a
replica shard can never advance its local checkpoint. As such operations
are just normal course of business for a replica shard, these operations
should be treated as if they completed successfully. This commit adds
these operations to the translog.
Relates #22626
For certain situations, end-users need the base path for Elasticsearch
logs. Exposing this as a property is better than hard-coding the path
into the logging configuration file as otherwise the logging
configuration file could easily diverge from the Elasticsearch
configuration file. Additionally, Elasticsearch will only have
permissions to write to the log directory configured in the
Elasticsearch configuration file. This commit adds a property that
exposes this base path.
One use-case for this is configuring a rollover strategy to retain logs
for a certain period of time. As such, we add an example of this to the
documentation.
Additionally, we expose the property es.logs.cluster_name as this is
used as the name of the log files in the default configuration.
Finally, we expose es.logs.node_name in cases where node.name is
explicitly set in case users want to include the node name as part of
the name of the log files.
Relates #22625
When logger.level is set, we end up configuring a logger named "level"
because we look for all settings of the form "logger\..+" as configuring
a logger. Yet, logger.level is special and is meant to only configure
the default logging level. This commit causes is to avoid not
configuring a logger named level.
Relates #22624
The IndexingOperationListener interface did not provide any
information about the shard id when a document was indexed.
This commit adds the shard id as the first parameter to all methods
in the IndexingOperationListener.
This commit is a simple cleanup of the code related to cgroup stats:
- reduce visibility of a method
- remove an unneeded logger guard
- cleanup the formatting of comments
TransportInterceptors are commonly used to enrich requests with headers etc.
which requires access the the thread context. This is not always easily possible
since threadpools are hard to access for instance if the interceptor is used on a transport client.
This commit passes on the thread context to all the interceptors for further consumption.
Closes#22585
Deleting indices is an important event in a cluster and as such should
be logged at the info level. This commit changes the logging level on
index deletion to the info level.
Relates #22627
We have made the security manager non-optional, but the Javadocs for
Security.java imply that it still is. This commit fixes this issue.
Relates #16176
ClusterService and TransportService expect the local discovery node to be set
before they are started but this requires manual interaction and is error prone since
to work absolutely correct they should share the same instance (same ephemeral ID).
TransportService also has 2 modes of operation, mainly realted to transport client vs. internal
to a node. This change removes the mode where we don't maintain a local node and uses a dummy local
node in the transport client since we don't bind to any port in such a case.
Local discovery node instances are now managed by the node itself and only suppliers and factories that allow
creation only once are passed to TransportService and ClusterService.
There was still small race in MockTcpTransport where channesl that are concurrently
closing are not yet removed from the reference tracking causing tests to fail. Compared to
the other races before this is a rather small windown and requires very very short test durations.
```h
$ bin/elasticsearch-keystore create
Created elasticsearch keystore in /Users/dpilato/Documents/Elasticsearch/apps/elasticsearch/elasticsearch-6.0.0-alpha1/config
$ bin/elasticsearch-keystore add
Enter value for null: xyz
Exception in thread "main" java.lang.NullPointerException: invalid null input
at java.security.KeyStore.setEntry(KeyStore.java:1552)
at org.elasticsearch.common.settings.KeyStoreWrapper.setString(KeyStoreWrapper.java:264)
at org.elasticsearch.common.settings.AddStringKeyStoreCommand.execute(AddStringKeyStoreCommand.java:83)
at org.elasticsearch.cli.EnvironmentAwareCommand.execute(EnvironmentAwareCommand.java:58)
at org.elasticsearch.cli.Command.mainWithoutErrorHandling(Command.java:122)
at org.elasticsearch.cli.MultiCommand.execute(MultiCommand.java:69)
at org.elasticsearch.cli.Command.mainWithoutErrorHandling(Command.java:122)
at org.elasticsearch.cli.Command.main(Command.java:88)
at org.elasticsearch.common.settings.KeyStoreCli.main(KeyStoreCli.java:39)
```
Today there are several races / holes in TcpTransport and MockTcpTransport
that can allow connections to be opened and remain unclosed while the actual
transport implementation is closed. A recently added assertions in #22554 exposes
these problems. This commit fixes several issues related to missed locks or channel
creations outside of a lock not checking if the resource is still open.
This change disables the _all meta field by default.
Now that we have the "all-fields" method of query execution, we can save both
indexing time and disk space by disabling it.
_all can no longer be configured for indices created after 6.0.
Relates to #20925 and #21341Resolves#19784
TcpTransport has an actual mechanism to stop resources in subclasses.
Instead of overriding `doStop` subclasses should override `stopInternal`
that is executed under the connection lock guaranteeing that there is no
concurrency etc.
Relates to #22554
* Settings: Make s3 repository sensitive settings use secure settings
This change converts repository-s3 to use the new secure settings. In
order to support the multiple ways we allow aws creds to be configured,
it also moves the main methods for the keystore wrapper into a
SecureSettings interface, in order to allow settings prefixing to work.
* Fix Translog.Delete serialization for sequence numbers
Translog.Delete used `.writeVLong` instead of `.writeLong` for the sequence
number and primary term (and their respective "read" variants). This could lead
to issues where a 5.x node sent a translog operation with a negative sequence
number (-2 for unassigned seq no) that tripped an assertion serializing a
negative number and causing ES to exit.
Adds a unit test for serialization and a mixed-cluster REST test, since that was
how this was originally caught.
* Use more realistic values for random seqNum and primary term
* Add comment with TODO for removal in 7.0
* Change comment into an assert
The low level TCP handshake can cause channel / connection leaks if it's interrupted
since the caller doesn't close the channel / connection if the handshake was not successful.
This commit fixes the channel leak and adds general test infrastructure to detect channel leaks
in the future.
Instead of `search.remote.seeds.${clustername}` we now specify the seeds as:
`search.remote.${clustername}.seeds` which is a real list setting compared to an unvalidated
group setting before.
Today affix settings are not dynamic since it's required to know
it's namespace in order to pull a concrete setting from it. This is not possible
in practice since the namespaces are dynamic by design. This change allows to register
a specialized settings consumer that consumes the namespace and the actual value if
a setting gets updated.
Moves fetching the local node id into `NodeClient` which is a
fairly useful place to put it so you can generate task ids from
`NodeClient#executeLocally`.
This commit adds the parsing fromXContent() methods to the IndexResponse class. The method is based on a ObjectParser because it is easier to use when parsing parent abstract classes like DocWriteResponse.
It also changes the ReplicationResponse.ShardInfo so that it now implements ToXContentObject. This way, the ShardInfo.fromXContent() method can be used by the IndexResponse's ObjectParser.
Previously, we removed all unneeded backward compatibility logic
from the BlobStoreRepository because 6.0 does not need to support
2.x snapshot formats. During the process of removing this backward
compatibility logic, some code was leftover that is no longer
necessary. This commit removes all the remaining unnecessary
backwards compatibility code in BlobStoreRepository.
It is no longer needed. It used to contain a lot of strings
used by serialization but those have since been removed. Now
it is just another thing to pass around that we don't really
need.
Affix settings are useful to namespace a certain setting. Yet, affix settings
must be specialized for their concrete type which causes lot of code duplication.
This commit allows to reuse an existing setting with and affix setting as soon as
a concrete key is available.
One needs to close the higher level objects (like UnicastZenPing) before closing the transport service. The latter can throw assertions w.r.t open connections
This adds methods to parse InternalSearchHit and InternalSearchHits from their
xContent representation. Most of the information in the original object is
preserved when rendering the object to xContent and then parsing it back.
However, some pieces of information are lost which we currently cannot parse
back from the rest response, most notably:
* the "match" property of the lucene explanation is not rendered in the
"_explain" section and cannot be reconstructed on the client side
* the original "shard" information (SearchShardTarget) is only rendered if the
"explanation" is also set, also we loose the indexUUID of the contained
ShardId because we don't write it out. As a replacement we can use
ClusterState.UNKNOWN_UUID on the receiving side
The NodeConnectionsService currently determines which nodes to connect to / disconnect from by inspecting cluster state changes and connecting to added nodes / disconnecting from removed nodes. When a master steps down (for example due to another master-eligible node shutting down which brings the number of master-eligible nodes below minimum_master_master), and the connection to other existing nodes was dropped while pinging, however, the connection to these nodes is not re-established while publishing the first cluster state that establishes the node as master.
This commit changes the NodeConnectionsService connect / disconnect logic to always rely on the state that is to be / was published, looking not only at the added / removed nodes, but validating that exactly all nodes that are currently registered in NodeConnectionsService are connected (corresponds to a NOOP if the node is already connected).
The document in the randomized GetResult can exist with no source (like if the _source was disabled in mappings), that's why the test should not always expect a non null source when the doc exists.
* Promote longs to doubles when a terms agg mixes decimal and non-decimal number
This change makes the terms aggregation work when the buckets coming from different indices are a mix of decimal numbers and non-decimal numbers. In this case non-decimal number (longs) are promoted to decimal (double) which can result in a loss of precision for big numbers.
Fixes#22232
There is a bug in the error message that is thrown if the number of docs differs between the source and target shards when recovering a shard with a syncId. The source and target doc counts are swapped around.
Closes#21893
Removes `AggregatorParsers`, replacing all of its functionality with
`XContentParser#namedObject`.
This is the third bit of payoff from #22003, one less thing to pass
around the entire application.
The test ping and waited for the ping results to be returned but since we first return the result and then close temporary connections, assertions are tripped that expects all connections to close by end of test .
Closes#22497
This commit checks for a null BytesReference as the value for `source`
in GetResult#sourceRef and simply returns null. Previously this would
have resulted in a NPE. While this does seem internal at first glance, it can affect
user code as a GetResponse could trigger this when the document is missing.
Additionally, the CompressorFactory#uncompressIfNeeded now requires a
non-null argument.
The recovery process started during primary relocation of shadow replicas accesses the engine on the source shard after it's been closed, which results in the source shard failing itself.
Right now closing a shard looks like it strands refresh listeners,
causing tests like
`delete/50_refresh/refresh=wait_for waits until changes are visible in search`
to fail. Here is a build that fails:
https://elasticsearch-ci.elastic.co/job/elastic+elasticsearch+multi_cluster_search+multijob-darwin-compatibility/4/console
This attempts to fix the problem by implements `Closeable` on
`RefreshListeners` and rejecting listeners when closed. More importantly
the act of closing the instance flushes all pending listeners
so we shouldn't have any stranded listeners on close.
Because it was needed for testing, this also adds the number of
pending listeners to the `CommonStats` object and all API to which
that flows: `_cat/nodes`, `_cat/indices`, `_cat/shards`, and
`_nodes/stats`.
In pre 2.x versions, if the repository was set to compress snapshots,
then snapshots would be compressed with the LZF algorithm. In 5.x,
Elasticsearch no longer supports the LZF compression algorithm. This
presents an issue when retrieving snapshots in a repository or upgrading
repository data to the 5.x version, because Elasticsearch throws an
exception when it tries to read the snapshot metadata because it was
compressed using LZF.
This commit gracefully handles the situation by introducing a new
incompatible-snapshots blob to the repository. For any pre-2.x snapshot
that cannot be read, that snapshot is removed from the list of active
snapshots, because the snapshot could not be restored anyway. Instead,
the snapshot is recorded in the incompatible-snapshots blob. When
listing snapshots, both active snapshots and incompatible snapshots will
be listed, with incompatible snapshots showing a `INCOMPATIBLE` state.
Any attempt to restore an incompatible snapshot will result in an
exception.
`ToXContentObject` extends `ToXContent` without adding new methods to it, while allowing to mark classes that output complete xcontent objects to distinguish them from classes that require starting and ending an anonymous object externally.
Ideally ToXContent would be renamed to ToXContentFragment, but that would be a huge change in our codebase, hence we simply document the fact that toXContent outputs fragments with no guarantees that the output is valid per se without an external ancestor.
Relates to #16347
This is related to #22116. A logIfNecessary() call makes a call to
NetworkInterface.getInterfaceAddresses() requiring SocketPermission
connect privileges. By moving this to bootstrap the logging call can be
made before installing the SecurityManager.
Today when an index is shrunk the version information is not carried over
from the source to the target index. This can cause major issues like mapping
incompatibilities for instance if an index from a previous major version is shrunk.
This commit ensures that all version information from the soruce index is preserved
when a shrunk index is created.
Closes#22373
ParseFieldMatcher as well as ParseFieldMatcherSupplier will be soon removed, hence the ObjectParser's context doesn't need to be a ParseFieldMatcherSupplier anymore. That will allow to remove ParseFieldMatcherSupplier's implementations, little by little.
The test currently checks that the recovering shard is not failed when it is not a primary relocation that has moved past the finalization step.
Checking if it has moved past that step is done by intercepting the request between the replication source and the target and checking if it has seen
then WAIT_FOR_CLUSTERSTATE action as this is the next action that is called after finalization. This action can, however, occur only after the shard was
already failed, and thus trip the assertion. This commit changes the check to look out for the FINALIZE action, independently of whether it succeeded or not.
#22325 changed the recovery retry logic to use unique recovery ids. The change also introduced an issue, however, which made it possible for the shard store to be closed under CancellableThreads, triggering assertions in the node locking logic. This commit limits the use of CancellableThreads only to the part where we wait on the old recovery target to be closed.
Today we execute the low level handshake on the TCP layer in #connectToNode.
If #openConnection is used directly, which is truly expert, no handshake is executed
which allows connecting to nodes that are not necessarily compatible. This change
moves the handshake to #openConnection to prevent bypassing this logic.
Previously, we could run into a situation where attempting to delete an
index due to a cluster state update would cause an unhandled exception
to bubble up to the ClusterService and cause the cluster state applier
to fail. The result of this situation is that the cluster state never
gets updated on the ClusterService because the exception happens before
all cluster state appliers have completed and the ClusterService only
updates the cluster state once all cluster state appliers have
successfully completed.
All other methods on IndicesService properly handle all exceptions and
not just IOExceptions, but there were two instances with respect to
index deletion where only IOExceptions where handled by the
IndicesService. If any other exception occurred during these delete
operations, the exception would be bubbled up to the ClusterService,
causing the aforementioned issues.
This commit ensures all methods in IndicesService properly capture all
types of Exceptions, so that the ClusterService manages to update the
cluster state, even in the presence of shard creation/deletion failures.
Note that the lack of updating the cluster state in the presence of such
exceptions can have many unintended consequences, one of them being
the tripping of the assertion in IndicesClusterStateService#removeUnallocatedIndices
where the assumption is that if there is an IndexService to remove with
an unassigned shard, then the index must exist in the cluster state, but if
the cluster state was never updated due to the aforementioned exceptions,
then the cluster state will not have the index in question.
Currently `geo_point` and `geo_shape` field are treated as `text` field by the field stats API and we
try to extract the min/max values with MultiFields.getTerms.
This is ok in master because a `geo_point` field is always a Point field but it can cause problem in 5.x (and 2.x) because the legacy
`geo_point` are indexed as terms.
As a result the min and max are extracted and then printed in the FieldStats output using BytesRef.utf8ToString
which can throw an IndexOutOfBoundException since it's not valid UTF8 strings.
This change ensure that we never try to extract min/max information from a `geo_point` field.
It does not add a new type for geo points in the fieldstats API so we'll continue to use `text` for this kind of field.
This PR is targeted to master even though we could only commit this change to 5.x. I think it's cleaner to have it in master too before we make any decision on
https://github.com/elastic/elasticsearch/pull/21947.
Fixes#22384
This commit cleans up the comments in IndexShard related to sequence numbers, making
them uniform in their formatting and taking advantage of the line-length
limit of 140 characters.
This commit cleans up the comments in GlobalCheckpointService, making
them uniform in their formatting and taking advantage of the line-length
limit of 140 characters.
This commit cleans up the comments in SequenceNumbersService, making
them uniform in their formatting and taking advantage of the line-length
limit of 140 characters.
this commit adds full support for proxy nodes on the search layer.
This allows to connection only to a small set of nodes on a remote cluster
to exectue the search. The nodes will proxy the request to the correct node in the
cluster while the coordinting node doesn't need to be connected to the target node.
This commit cleans up the comments in LocalCheckpointService, making
them uniform in their formatting and taking advantage of the line-length
limit of 140 characters.
After deprecating getters and setters and the query DSL parameter in 5.x,
support for `minimum_number_should_match` can be removed entirely. Also
consolidated comments with the ones on 5.x branch and added an entry to the
migration docs.
Currently we have getters an setters for both "minimumNumberShouldMatch" and
"minimumShouldMatch", which both access the same internal value
(minimumShouldMatch). Since we only document the `minimum_should_match`
parameter for the query DSL, I think we can deprecate the other getters and
setters for 5.x and remove with 6.0, also deprecating the
`minimum_number_should_match` query DSL parameter.
This PR completes the refactoring of the cluster allocation explain API and improves it in the following two high-level ways:
1. The explain API now uses the same allocators that the AllocationService uses to make shard allocation decisions. Prior to this PR, the explain API would run the deciders against each node for the shard in question, but this was not executed on the same code path as the allocators, and many of the scenarios in shard allocation were not captured due to not executing through the same code paths as the allocators.
2. The APIs have changed, both on the Java and JSON level, to accurately capture the decisions made by the system. The APIs also now report on shard moving and rebalancing decisions, whereas the previous API did not report decisions for moving shards which cannot remain on their current node or rebalancing shards to form a more balanced cluster.
Note: this change affects plugin developers who may have a custom implementation of the ShardsAllocator interface. The method weighShards has been removed and no longer has any utility. In order to support the new explain API, however, a custom implementation of ShardsAllocator must now implement ShardAllocationDecision decideShardAllocation(ShardRouting shard, RoutingAllocation allocation) which provides a decision and explanation for allocating a single shard. For implementations that do not support explaining a single shard allocation via the cluster allocation explain API, this method can simply return an UnsupportedOperationException.
In #22313 we added a check that prevents the SnapshotDeletionsInProgress custom cluster state objects from being sent to older elasticsearch nodes. This commits make this check generic and available to other cluster state custom objects if needed.
Unless the dynamic templates define an explicit format in the mapping
definition: in that case the explicit mapping should have precedence.
Closes#9410
This adds a new `normalizer` property to `keyword` fields that pre-processes the
field value prior to indexing, but without altering the `_source`. Note that
only the normalization components that work on a per-character basis are
applied, so for instance stemming filters will be ignored while lowercasing or
ascii folding will be applied.
Closes#18064
Resetting a recovery consists of resetting the old recovery target and replacing it by a new recovery target object. This is done on the Cancellable threads of
the new recovery target. If the new recovery target is already cancelled before or while this happens, for example due to shard closing or recovery source
changing, we have to make sure that the old recovery target object frees all shard resources.
Relates to #22325
Recoveries are tracked on the target node using RecoveryTarget objects that are kept in a RecoveriesCollection. Each recovery has a unique id that is communicated from the recovery target to the source so that it can call back to the target and execute actions using the right recovery context. In case of a network disconnect, recoveries are retried. At the moment, the same recovery id is reused for the restarted recovery. This can lead to confusion though if the disconnect is unilateral and the recovery source continues with the recovery process. If the target reuses the same recovery id while doing a second attempt, there might be two concurrent recoveries running on the source for the same target.
This commit changes the recovery retry process to use a fresh recovery id. It also waits for the first recovery attempt to be fully finished (all resources locally freed) to further prevent concurrent access to the shard. Finally, in case of primary relocation, it also fails a second recovery attempt if the first attempt moved past the finalization step, as the relocation source can then be moved to RELOCATED state and start indexing as primary into the target shard (see TransportReplicationAction). Resetting the target shard in this state could mean that indexing is halted until the recovery retry attempt is completed and could also destroy existing documents indexed and acknowledged before the reset.
Relates to #22043
`scaled_float` should be used as DOUBLE in aggregations but currently they are used as LONG.
This change fixes this issue and adds a simple it test for it.
Fixes#22350
Before, snapshot/restore would synchronize all operations on the cluster
state except for deleting snapshots. This meant that only one
snapshot/restore operation would be allowed in the cluster at any given
time, except for deletions - there could be two or more snapshot
deletions running at the same time, or a deletion could be running,
unbeknowest to the rest of the cluster, and thus a snapshot or restore
would be allowed at the same time as the snapshot deletion was still in
progress. This could cause any number of synchronization issues,
including the situation where a snapshot that was deleted could reappear
in the index-N file, even though its data was no longer present in the
repository.
This commit introduces a new custom type to the cluster state to
represent deletions in progress. Now, another deletion cannot start if
a deletion is currently in progress. Similarily, a snapshot or restore
cannot be started if a deletion is currently in progress. In each case,
if attempting to run another snapshot/restore operation while a deletion
is in progress, a ConcurrentSnapshotExecutionException will be thrown.
This is the same exception thrown if trying to snapshot while another
snapshot is in progress, or restore while a snapshot is in progress.
Closes#19957
This commit fixes an issue with IndexShardTests#testDocStats when the
number of deleted docs is equal to the number of docs. In this case,
Luence will remove the underlying segment tripping an assertion on the
number of deleted docs.
Today we try to pull stats from index writer but we do not get a
consistent view of stats. Under heavy indexing, this inconsistency can
be very skewed indeed. In particular, it can lead to the number of
deleted docs being reported as negative and this leads to serialization
issues. Instead, we should provide a consistent view of the stats by
using an index reader.
Relates #22317
Not doing this made it difficult to establish a happens before relationship between connecting to a node and adding a listeners. Causing test code like this to fail sproadically:
```
// connection to reuse
handleA.transportService.connectToNode(handleB.node);
// install a listener to check that no new connections are made
handleA.transportService.addConnectionListener(new TransportConnectionListener() {
@Override
public void onConnectionOpened(DiscoveryNode node) {
fail("should not open any connections. got [" + node + "]");
}
});
```
relates to #22277
This commit factors out the cluster state update tasks that are published (ClusterStateUpdateTask) from those that are not (LocalClusterUpdateTask), serving as a basis for future refactorings to separate the publishing mechanism out of ClusterService.
When starting a standalone cluster, we do not able assertions. This is
problematic because it means that we miss opportunities to catch
bugs. This commit enables assertions for standalone integration tests,
and fixes a couple bugs that were uncovered by enabling these.
Relates #22334
This change is the first towards providing the ability to store
sensitive settings in elasticsearch. It adds the
`elasticsearch-keystore` tool, which allows managing a java keystore.
The keystore is loaded upon node startup in Elasticsearch, and used by
the Setting infrastructure when a setting is configured as secure.
There are a lot of caveats to this PR. The most important is it only
provides the tool and setting infrastructure for secure strings. It does
not yet provide for keystore passwords, keypairs, certificates, or even
convert any existing string settings to secure string settings. Those
will all come in follow up PRs. But this PR was already too big, so this
at least gets a basic version of the infrastructure in.
The two main things to look at. The first is the `SecureSetting` class,
which extends `Setting`, but removes the assumption for the raw value of the
setting to be a string. SecureSetting provides, for now, a single
helper, `stringSetting()` to create a SecureSetting which will return a
SecureString (which is like String, but is closeable, so that the
underlying character array can be cleared). The second is the
`KeyStoreWrapper` class, which wraps the java `KeyStore` to provide a
simpler api (we do not need the entire keystore api) and also extend
the serialized format to add metadata needed for loading the keystore
with no assumptions about keystore type (so that we can change this in
the future) as well as whether the keystore has a password (so that we
can know whether prompting is necessary when we add support for keystore
passwords).
We don't *want* to use negative numbers with `writeVLong`
so throw an exception when we try. On the other
hand unforeseen bugs might cause us to write negative numbers (some versions of Elasticsearch don't have the exception, only an assertion)
so this fixes `readVLong` so that instead of reading a wrong
value and corrupting the stream it reads the negative value.
Optimistically check for `tag` of an unknown processor for better tracking of which
processor declaration is to blame in an invalid configuration.
Closes#21429.
* Remove a checked exception, replacing it with `ParsingException`.
* Remove all Parser classes for the yaml sections, replacing them with static methods.
* Remove `ClientYamlTestFragmentParser`. Isn't used any more.
* Remove `ClientYamlTestSuiteParseContext`, replacing it with some static utility methods.
I did not rewrite the parsers using `ObjectParser` because I don't think it is worth it right now.
Currently we only apply date detection on strings that contain either `:`, `-`
or `/`. This commit inverses the heuristic in order to only apply date detection
on strings that are not parseable as a number, so that more date formats can be
used as dynamic dates formats.
Closes#1694
Today we only expose `value_type` in scriptable aggregations, however it is
also useful with unmapped fields. I suspect we never noticed because
`value_type` was not documented (fixed) and most aggregations are scriptable.
Closes#20163
`ShardCoreKeyMap.add` is called on each segment for all search requests, which
means it might become a bottleneck under a cocurrent load of cheap search
requests since this method acquires a mutex. This change proposes to use a
`ConcurrentHashMap` which allows to only take the mutex in the case that the
`LeafReader` has never been seen before.
This adds test classes that can be used to test the wire serialisation and (optionally) the XContent serialisation of objects that implement Streamable/Writeable and ToXContent.
These test classes will enable classes sich as InternalAggregation (or at least its implementations) to be tested in a consistent way when is comes to testing serialisation.
As the translog evolves towards a full operations log as part of the
sequence numbers push, there is a need for the translog to be able to
represent operations for which a sequence number was assigned, but the
operation did not mutate the index. Examples of how this can arise are
operations that fail after the sequence number is assigned, and gaps in
this history that arise when an operation is assigned a sequence number
but the operation never completed (e.g., a node crash). It is important
that these operations appear in the history so that they can be
replicated and replayed during recovery as otherwise the history will be
incomplete and local checkpoints will not be able to advance. This
commit introduces a no-op to the translog to set the stage for these
efforts.
Relates #22291
Today if an older version of a plugin exists, we fail to notify the user
with a helpful error message. This happens because during plugin
verification, we attempt to read the plugin descriptors for all existing
plugins. When an older version of a plugin is sitting on disk, we will
attempt to read this old plugin descriptor and fail due to a version
mismatch. This leads to an unhelpful error message. Instead, we should
check for existence of the plugin as part of the verification phase, but
before attempting to read plugin descriptors for existing plugins. This
enables us to provide a helpful error message to the user.
Relates #22305
The deprecation warning gives now the same message as 5.x. The deprecation warning was previously removed, but given that we are still lenient with old indices we should still output the warning.
Our `float`/`double` fields generally assume that `-0` compares less than `+0`,
except when bounds are exclusive: an exclusive lower bound on `-0` excludes
`+0` and an exclusive upper bound on `+0` excludes `-0`.
Closes#22167
The way aggregations on scripts work is by hiding scripts behind the same API
that we use for regular fields. However, there is no native support for boolean
fields, those need to be exposed as integers, with `0` standing for `false` and
`1` for true.
Relates #20941
The `UnicastZenPing` shows it's age and is the result of many small changes. The current state of affairs is confusing and is hard to reason about. This PR cleans it up (while following the same original intentions). Highlights of the changes are:
1) Clear 3 round flow - no interleaving of scheduling.
2) The previous implementation did a best effort attempt to wait for ongoing pings to be sent and completed. The pings were guaranteed to complete because each used the total ping duration as a timeout. This did make it hard to reason about the total ping duration and the flow of the code. All of this is removed now and ping should just complete within the given duration or not be counted (note that it was very handy for testing, but I move the needed sync logic to the test).
3) Because of (2) the pinging scheduling changed a bit, to give a chance for the last round to complete. We now ping at the beginning, 1/3 and 2/3 of the duration.
4) To offset for (3) a bit, incoming ping requests are now added to on going ping collections.
5) UnicastZenPing never establishes full blown connections (but does reuse them if there). Relates to #22120
6) Discovery host providers are only used once per pinging round. Closes#21739
7) Usage of the ability to open a connection without connecting to a node ( #22194 ) and shorter connection timeouts helps with connections piling up. Closes#19370
8) Beefed up testing and sped them up.
9) removed light profile from production code
This adds fromXContent method and unit test for sort values that are part of
InternalSearchHit. In order to centralize serialisation and xContent parsing and
rendering code, move all relevant parts to a new class which can be unit tested
much better in isolation.This is part of the preparation for parsing search
responses on the client side.
Sending a request is not a good indicator as it doesn't mean it's processed yet. Instead we should use one of the first request from source to target.
This caused the cluster state block to be added to early , blocking the recovery it self
The allocation decider explanation messages where improved in #21771 to
include the specific Elasticsearch setting that contributed to the
decision taken by the decider. This commit improves upon the
explanation message output by including whether the setting was an index
level setting or a cluster level setting. This will further help the
user understand and locate the setting that is the cause of shards
remaining unassigned or remaining on their current node.
Introduces `XContentParser#namedObject which works a little like
`StreamInput#readNamedWriteable`: on startup components register
parsers under names and a superclass. At runtime we look up the
parser and call it to parse the object.
Right now the parsers take a context object they use to help with
the parsing but I hope to be able to eliminate the need for this
context as most what it is used for at this point is to move
around parser registries which should be replaced by this method
eventually. I make no effort to do so in this PR because it is
big enough already. This is meant to the a start down a road that
allows us to remove classes like `QueryParseContext`,
`AggregatorParsers`, `IndicesQueriesRegistry`, and
`ParseFieldRegistry`.
The goal here is to reduce the amount of plumbing required to
allow parsing pluggable things. With this you don't have to pass
registries all over the place. Instead you must pass a super
registry to fewer places and use it to wrap the reader. This is
the same tradeoff that we use for NamedWriteable and it allows
much, much simpler binary serialization. We think we want that
same thing for xcontent serialization.
The only parsing actually converted to this method is parsing
`ScoreFunctions` inside of `FunctionScoreQuery`. I chose this
because it is relatively self contained.
ClusterStateObserver is a utility class that simplifies interacting with the cluster state in cases where an action takes a decision based on the current cluster state but may want to wait for a new state and retry upon failure. The ClusterStateObserver implements its functionality by keeping a reference to the last cluster state that it observed. When a new ClusterStateObserver is created, it samples a cluster state from the cluster service which is subsequently used for change detection. If actions take a long time to process, however, the cluster observer can reference very old cluster states. Due to cluster observers being created very frequently and cluster states being potentially large the referenced cluster states can waste a lot of heap space. A specific example where this can make a node go out of memory is given in point 2 of issue #21568: The action listener in TransportMasterNodeAction.AsyncSingleAction has a ClusterStateObserver to coordinate the retry mechanism if the action on the master node fails due to the node not being master anymore. The ClusterStateObserver in AsyncSingleAction keeps a reference to the full cluster state when the action was initiated. If the pending tasks queue grows quite large and has older items in it lots of cluster states can possibly be referenced.
This commit changes the ClusterStateObserver to hold only onto the part of the cluster state that's needed for change detection.
This changes the class from extending the abstract class to implementing the
ToXContent interface only. The former could lead to unexpected behaviour when
trying to display the object, since the "toString()" method inherited from
ToXContentToBytes would create an error message because the SuggestionBuilders
toXContent() methods don't render complete json objects.
* Internal: Refactor SettingCommand into EnvironmentAwareCommand
This change renames and changes the behavior of SettingCommand to have
its primary method take in a fully initialized Environment for
elasticsearch instead of just a map of settings. All of the subclasses
of SettingCommand already did this at some point, so this just removes
duplication.
We are currenlty checking that no deprecation warnings are emitted in our query tests. That can be moved to ESTestCase (disabled in ESIntegTestCase) as it allows us to easily catch where our tests use deprecated features and assert on the expected warnings.
We return deprecation warnings as response headers, besides logging them. Strict parsing mode stayed around, but was only used in query tests, though we also introduced checks for deprecation warnings there that don't need strict parsing anymore (see #20993).
We can then safely remove support for strict parsing mode. The final goal is to remove the ParseFieldMatcher class, but there are many many users of it. This commit prepares the field for the removal, by deprecating ParseFieldMatcher and making it effectively not needed. Strict parsing is removed from ParseFieldMatcher, and strict parsing is replaced in tests where needed with deprecation warnings checks.
Note that the setting to enable strict parsing was never ported to the new settings infra hance it cannot be set in production. It is really only used in our own tests.
Relates to #19552
Rename the method to assertToXContentEquivalent to highlight that it's tailored to ToXContent comparisons.
Rather than parsing into a map and replacing byte[] in both those maps, add custom equality assertions that recursively walk maps and lists and call Arrays.equals whenever a byte[] is encountered.
Moved field values `toXContent` logic to `GetField` (from `GetResult`), which outputs its own fields, and can also parse them now. Also added `fromXContent` to `GetResult` and `GetResponse`.
The start object and end object for `GetResponse` output have been moved to `GetResult#toXContent`, from the corresponding rest action. This makes it possible to have `toXContent` and `fromXContent` completely symmetric, as parsing requires looping till an end object is found which is weird when the corresponding `toXContent` doesn't print that out.
This also introduces the foundation for testing retrieval of _source and stored field values.
This commit makes mapping updates atomic when multiple types in an index are updated. Mappings for an index are now applied in a single atomic operation, which also allows to optimize some of the cross-type updates and checks.
Subclasses of TransportReplicationAction can currently chose to implement block levels for which the request will be blocked.
- Refresh/Flush was using the block level METADATA_WRITE although they don't operate at the cluster meta data level (but more like shard level meta data which is not represented in the block levels). Their level has been changed to null so that they can operate freely in the presence of blocks.
- GlobChkptSync was using WRITE although it does not make any changes to the actual documents of a shard. The level has been changed to null so that it can operate freely in the presence of blocks.
The commit also adds a check for closed indices in TRA so that the right exception is thrown if refresh/flush/checkpoint syncing is attempted on a closed index (before it was throwing an IndexNotFoundException, now it's throwing IndexClosedException).
Sequence BWC logic consists of two elements:
1) Wire level BWC using stream versions.
2) A changed to the global checkpoint maintenance semantics.
For the sequence number infra to work with a mixed version clusters, we have to consider situation where the primary is on an old node and replicas are on new ones (i.e., the replicas will receive operations without seq#) and also the reverse (i.e., the primary sends operations to a replica but the replica can't process the seq# and respond with local checkpoint). An new primary with an old replica is a rare because we do not allow a replica to recover from a new primary. However, it can occur if the old primary failed and a new replica was promoted or during primary relocation where the source primary is treated as a replica until the master starts the target.
1) Old Primary & New Replica - this case is easy as is taken care of by the wire level BWC. All incoming requests will have their seq# set to `UNASSIGNED_SEQ_NO`, which doesn't confuse the local checkpoint logic (keeping it at `NO_OPS_PERFORMED`)
2) New Primary & Old replica - this one is trickier as the global checkpoint service currently takes all in sync replicas into consideration for the global checkpoint calculation. In order to deal with old replicas, we change the semantics to say all *new node* in sync replicas. That means the replicas on old nodes don't count for the global checkpointing. In this state the seq# infra is not fully operational (you can't search on it, because copies may miss it) but it is maintained on shards that can support it. The old replicas will have to go through a file based recovery at some point and will get the seq# information at that point. There is still an edge case where a new primary fails and an old replica takes over. I'lll discuss this one with @ywelsch as I prefer to avoid it completely.
This PR also re-enables the BWC tests which were disabled. As such it had to fix any BWC issue that had crept in. Most notably an issue with the removal of the `timestamp` field in #21670.
The commit also includes a fix for the default value of the seq number field in replicated write requests (it was 0 but should be -2), that surface some other minor bugs which are fixed as well.
Last - I added some debugging tools like more sane node names and forcing replication request to implement a `toString`
This commit exposes public getters for the aggregations in
AggregatorFactories.Builder. The reason is that it allows to
parse the aggregation object from elsewhere (e.g. a plugin) and then
be able to get the aggregation builders in order to set them in
a SearchSourceBuilder.
The alternative would have been to expose a setter for the
AggregatorFactories.Builder object. But that would be making
the API a bit trappy.
Today if a settings object has many keys ie. if somebody specifies
a gazillion synonym in-line (arrays are keys ending with ordinals) operations like
`Settings#getByPrefix` have a linear runtime. This can cause index creations to be
very slow producing lots of garbage at the same time. Yet, `Settings#getByPrefix` is called
quite frequently by group settings etc. which can cause heavy load on the system.
While it's not recommended to have synonym lists with 25k entries in-line these use-cases should
not have such a large impact on the cluster / node. This change introduces a view-like map
that filters based on the prefixes referencing the actual source map instead of copying all values
over and over again. A benchmark that adds a single key with 25k random synonyms between 2 and 5 chars
takes 16 seconds to get the synonym prefix 200 times while the filtered view takes 4 ms for the 200 iterations.
This relates to https://discuss.elastic.co/t/200-cpu-elasticsearch-5-index-creation-very-slow-with-a-huge-synonyms-list/69052
In some cases, it might happen that the `_all` field gets a field type that is
not totally configured, and in particular lacks analyzers. This is due to the
fact that `AllFieldMapper.TypeParser.getDefault` uses `Defaults.FIELD_TYPE` as
a default field type, which does not have any analyzers configured since it
does not know about the default analyzers.
With this commit we enable the Jackson feature 'STRICT_DUPLICATE_DETECTION'
by default for all XContent types (not only JSON).
We have also changed the name of the system property to disable this feature
from `es.json.strict_duplicate_detection` to the now more appropriate name
`es.xcontent.strict_duplicate_detection`.
Relates elastic/elasticsearch#19614
Relates elastic/elasticsearch#22073
With this commit we change the data type of the 'TIMESTAMP'
meta-data field from a formatted date string to a plain
`java.util.Date` instance. The main reason for this change is
that our benchmarks have indicated that this contributes
significantly to the time spent in the ingest pipeline.
The overhead in terms of indexing throughput of the ingest
pipeline is about 15% and breaks down roughly as follows:
* 5% overhead caused by the conversion from `XContent` -> `Map`
* 5% overhead caused by the timestamp formatting
* 5% overhead caused by the conversion `Map` -> `XContent`
Relates #22074
In #22094 we introduce a test-only setting to simulate transport
impls that don't support handshakes. This commit implements the same logic
without a setting.
This commit touches addresses issues related to recovery and sequence numbers:
- A sequence number can be assigned and a Lucene commit created with a
maximum sequence number at least as large as that sequence number,
yet the operation corresponding to that sequence number can be
missing from both the Lucene commit and the translog. This means that
upon recovery the local checkpoint will be stuck at or below this
missing sequence number. To address this, we force the local
checkpoint to the maximum sequence number in the Lucene commit when
opening the engine. Note that there can still be gaps in the history
in the translog but we do not address those here.
- The global checkpoint is transferred to the target shard at the end
of peer recovery.
- Additionally, we reenable the relocation integration tests.
Lastly, this work uncovered some bugs in the assignment of sequence
numbers on replica operations:
- setting the sequence number on replica write requests was missing,
very likely introduced as a result of resolving merge conflicts
- handling operations that arrive out of order on a replica and have a
version conflict with a previous operation were never marked as
processed
Relates #22212
Some expert users like UnicastZenPing today establishes real connections to nodes during it's ping
phase that can be used by other parts of the system. Yet, this is potentially dangerous
and undesirable unless the nodes have been fully verified and should be connected to in the
case of a cluster state update or if we join a newly elected master. For use-cases like this, this change adds the infrastructure to manually handle connections that are not publicly available on the node ie. should not be managed by `Transport`/`TransportSerivce`
Some of our stats serialization code duplicates complicated seriazliation logic
or could use existing building blocks from StreamOutput/Input. This commit
cleans up some of the serialization code.
Today in the codebase we refer to seccomp everywhere instead of system
call filter even if we are not specifically referring to Linux. This
commit is a purely mechanical change to refer to system call filter
where appropriate instead of the general seccomp, and only leaves
seccomp in place when actually referring to the Linux implementation.
Relates #22243
We try to install a system call filter on various operating systems
(Linux, macOS, BSD, Solaris, and Windows) but the setting
(bootstrap.seccomp) to control this is named after the Linux
implementation (seccomp). This commit replaces this setting with
bootstrap.system_call_filter. For backwards compatibility reasons, we
fallback to bootstrap.seccomp and log a deprecation message if
bootstrap.seccomp is set. We intend to remove this fallback in
6.0.0. Note that now is the time to make this change it's likely that
most users are not making this setting anyway as prior to version 5.2.0
(currently unreleased) it was not necessary to configure anything to
enable a node to start up if the system call filter failed to install
(we marched on anyway) but starting in 5.2.0 it will be necessary in
this case.
Relates #22226
With recent changes to our parsing code we have drastically reduced the places where we auto-detect the content type from the input. The usage of these methods spread in our codebase for no reason, given that in most of the cases we know the content type upfront and we don't need any auto-detection mechanism. Deprecating these methods is a way to try and make sure that these methods are carefully used, and hopefully not introduced in newly written code.
We have yet to fix the REST layer to read the Content-Type header, which is the long term solution, but for now we just want to make sure that the usage of these methods doesn't spread any further.
Relates to #19388
In #20305, _suggest endpoint was deprecated
in favour of using _search endpoint. This
commit removes the dedicated _suggest endpoint
entirely from master.
Depending on how the connection is closed the `#onChannelClosed` callback
might be invoked more than once or the handler has been processed by the response
of the handshake already. This commit only notifies the handler if was removed from
the pending map.
Sends the `error_trace` parameter with all requests sent by the
yaml test framework, including the doc snippet tests. This can be
overridden by settings `error_trace: false`. While this drift's
core's handling of the yaml tests from the client's slightly this
should only be a problem for tests that rely on the default value,
both of which I've fixed by setting the value explicitly.
This also escapes `\n` and `\t` in the `Stash dump on failure` so
the `stack_trace` is more readable.
Also fixes `RestUpdateSettingsAction` to not think of the `error_trace`
parameter as a setting.
`ClusterService` is responsible of updating the cluster state on every node (as a response to an API call on the master and when non-masters receive a new state from the master). When a new cluster state is processed, it is made visible via the `ClusterService#state` method and is sent to series of listeners. Those listeners come in two flavours - one is to change the state of the node in response to the new cluster state (call these cluster state appliers), the other is to start a secondary process. Examples for the later include an indexing operation waiting for a shard to be started or a master node action waiting for a master to be elected.
The fact that we expose the state before applying it means that samplers of the cluster state had to worry about two things - working based on a stale CS and working based on a future, i.e., "being applied" CS. The `ClusterStateStatus` was used to allow distinguishing between the two. Working with a stale cluster state is not avoidable. How this PR changes things to make sure consumers don't need to worry about future CS, removing the need for the status and simplifying the waiting logic.
This change does come with a price as "cluster state appliers" can't sample the cluster state from `ClusterService` whenever they want as the cluster state isn't exposed yet. However, recent clean ups made this is situation easier and this PR takes the last steps to remove such sampling. This also helps clarify the "information flow" and helps component separation (and thus potential unit testing). It also adds an assertion that will trigger if the cluster state is sampled by such listeners.
Note that there are still many "appliers" that could be made a simpler, unrestricted "listener" but this can be done in smaller bits in the future. The commit also makes it clear what the `appliers` and what the `listeners` are by using dedicated interfaces.
Also, since I had to change the listener types I went ahead and changed the data structure for temporary/timeout listeners (used for the observer) so addition and removal won't be an O(n) operation.
Simon Willnauer
Nik Everett
Nicholas Knize
Christoph Büscher
Ali Beyad
Adrien Grand
Boaz Leskes
Martijn van Groningen
Yannick Welsch
Lee Hinman
Tim Brooks
Jason Tedor
Jim Ferenczi
Chris Buonocore
Areek Zillur
Jay Modi
Igor Motov
Tanguy Leroux
Jack Conradson
Ryan Ernst
Alex Bumbu
Colin Goodheart-Smithe
javanna
srgclr
Luca Cavanna
Alexander Reelsen
markharwood
David Pilato
Daniel Mitterdorfer
Daniel Mitterdorfer
Michael McCandless
Ke Li
Scott Somerville
Matt Weber
Lee Hinman
Simon Willnauer
henakamaMSFT
Tim Vernum
Tal Levy
Stéphane Campinas
Alexander Lin
Grzegorz Gajos
Dimitris Athanasiou
Masaru Hasegawa