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
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`
Jason Tedor
Boaz Leskes
Yannick Welsch
Nik Everett
Tal Levy
Stéphane Campinas
Martijn van Groningen
Colin Goodheart-Smithe
Adrien Grand
Luca Cavanna
Christoph Büscher
Ali Beyad
Tanguy Leroux
Ryan Ernst
Alexander Lin
Grzegorz Gajos
javanna