Local and global checkpoints currently do not correctly reflect what's persisted to disk. The issue is
that the local checkpoint is adapted as soon as an operation is processed (but not fsynced yet). This
leaves room for the history below the global checkpoint to still change in case of a crash. As we rely
on global checkpoints for CCR as well as operation-based recoveries, this has the risk of shard
copies / follower clusters going out of sync.
This commit required changing some core classes in the system:
- The LocalCheckpointTracker keeps track now not only of the information whether an operation has
been processed, but also whether that operation has been persisted to disk.
- TranslogWriter now keeps track of the sequence numbers that have not been fsynced yet. Once
they are fsynced, TranslogWriter notifies LocalCheckpointTracker of this.
- ReplicationTracker now keeps track of the persisted local and persisted global checkpoints of all
shard copies when in primary mode. The computed global checkpoint (which represents the
minimum of all persisted local checkpoints of all in-sync shard copies), which was previously stored
in the checkpoint entry for the local shard copy, has been moved to an extra field.
- The periodic global checkpoint sync now also takes async durability into account, where the local
checkpoints on shards only advance when the translog is asynchronously fsynced. This means that
the previous condition to detect inactivity (max sequence number is equal to global checkpoint) is
not sufficient anymore.
- The new index closing API does not work when combined with async durability. The shard
verification step is now requires an additional pre-flight step to fsync the translog, so that the main
verify shard step has the most up-to-date global checkpoint at disposition.
The RemoteClusterService should close the current
RemoteClusterConnection and should build it again if
the seeds are changed, similarly to what is done when
the ping interval or the compression settings are changed.
Closes#37799
The fact that SearchPhaseContext extends ActionListener makes it hard
to reason about when the original listener is notified and to trace
those calls. Also, the corresponding onFailure and onResponse were
only needed in two places, one each, where they can be replaced by a
more intuitive call, like sendSearchResponse for onResponse.
Today the fielddata for global ordinals re-creates docvalues readers of each segment
when building the iterator of a single segment. This is required because the lookup of
global ordinals needs to access the docvalues's TermsEnum of each segment to retrieve
the original terms. This also means that we need to create NxN (where N is the number of segment in the index) docvalues iterators
each time we want to collect global ordinal values. This wasn't an issue in previous versions since docvalues readers are stateless
before 6.0 so they are reused on each segment but now that docvalues are iterators we need to create a new instance each time
we want to access the values. In order to avoid creating too many iterators this change splits
the global ordinals fielddata in two classes, one that is used to cache a single instance per directory reader and one
that is created from the cached instance that can be used by a single consumer. The latter creates the TermsEnum of each segment
once and reuse them to create the segment's iterator. This prevents the creation of all TermsEnums each time we want to access
the value of a single segment, hence reducing the number of docvalues iterator to create to Nx2 (one iterator and one lookup per segment).
This commit removes some very old test logging annotations that appeared
to be added to investigate test failures that are long since closed. If
these are needed, they can be added back on a case-by-case basis with a
comment associating them to a test failure.
It's possible for the passed in `IndexMetaData` to be null (for
instance, cluster state passed in does not have the index in its
metadata) which in turn can cause a `NullPointerException` when
evaluating the conditions for an index. This commit adds null protection
and unit tests for this case.
Resolves#43296
While investigating memory consumption of deeply nested aggregations for #43091
the memory used to keep track of the doc ids and buckets in the BestBucketsDeferringCollector
showed up as one of the main contributor. In my tests half of the memory held in the
BestBucketsDeferringCollector is associated to segments that don't have matching docs
in the selected buckets. This is expected on fields that have a big cardinality since each
bucket can appear in very few segments. By allocating the builders lazily this change
reduces the memory consumption by a factor 2 (from 1GB to 512MB), hence reducing the
impact on gcs for these volatile allocations. This commit also switches the PackedLongValues.Builder
with a RoaringDocIdSet in order to handle very sparse buckets more efficiently.
I ran all my tests on the `geoname` rally track with the following query:
````
{
"size": 0,
"aggs": {
"country_population": {
"terms": {
"size": 100,
"field": "country_code.raw"
},
"aggs": {
"admin1_code": {
"terms": {
"size": 100,
"field": "admin1_code.raw"
},
"aggs": {
"admin2_code": {
"terms": {
"size": 100,
"field": "admin2_code.raw"
},
"aggs": {
"sum_population": {
"sum": {
"field": "population"
}
}
}
}
}
}
}
}
}
}
````
This PR is a backport a of #43214 from v8.0.0
A number of the aggregation base classes have an abstract doEquals() and doHashCode() (e.g. InternalAggregation.java, AbstractPipelineAggregationBuilder.java).
Theoretically this is so the sub-classes can add to the equals/hashCode and don't need to worry about calling super.equals(). In practice, it's mostly just confusing/inconsistent. And if there are more than two levels, we end up with situations like InternalMappedSignificantTerms which has to call super.doEquals() which defeats the point of having these overridable methods.
This PR removes the do versions and just use equals/hashCode ensuring the super when necessary.
* Follow up to #42109:
* Adjust test to only check that interface lookup by name works not actually lookup IPs which is brittle since virtual interfaces can be destroyed/created by Docker while the tests are running
Co-authored-by: Jason Tedor <jason@tedor.me>
* Return 0 for negative "free" and "total" memory reported by the OS
We've had a situation where the MX bean reported negative values for the
free memory of the OS, in those rare cases we want to return a value of
0 rather than blowing up later down the pipeline.
In the event that there is a serialization or creation error with regard
to memory use, this adds asserts so the failure will occur as soon as
possible and give us a better location for investigation.
Resolves#42157
* Fix test passing in invalid memory value
* Fix another test passing in invalid memory value
* Also change mem check in MachineLearning.machineMemoryFromStats
* Add background documentation for why we prevent negative return values
* Clarify comment a bit more
This PR reverts #35230.
Previously, we reply on soft-deletes to fill the mismatch between the
version map and the Lucene index. This is no longer needed after #43202
where we rebuild the version map when opening an engine. Moreover,
PrunePostingsMergePolicy can prune _id of soft-deleted documents out of
order; thus the lookup result including soft-deletes sometimes does not
return the latest version (although it's okay as we only use a valid
result in an engine).
With this change, we use only live documents in Lucene to resolve the
indexing strategy. This is perfectly safe since we keep all deleted
documents after the local checkpoint in the version map.
Closes#42979
Backport of: https://github.com/elastic/elasticsearch/pull/43222
This commit replaces usages of Streamable with Writeable for the
SingleShardRequest / TransportSingleShardAction classes and subclasses of
these classes.
Note that where possible response fields were made final and default
constructors were removed.
Relates to #34389
Today the `ClusterFormationFailureHelper` does not include the local node in
the list of nodes it claims to have discovered. This means that it sometimes
reports that it has not discovered a quorum when in fact it has. This commit
adds the local node to the set of discovered nodes.
Today we suppress election attempts on master-eligible nodes that are not in
the voting configuration. In fact this restriction is not necessary: any
master-eligible node can safely become master as long as it has a fresh enough
cluster state and can gather a quorum of votes. Moreover, this restriction is
sometimes undesirable: there may be a reason why we do not want any of the
nodes in the voting configuration to become master.
The reason for this restriction is as follows. If you want to shut the master
down then you might first exclude it from the voting configuration. When this
exclusion succeeds you might reasonably expect that a new master has been
elected, since the voting config exclusion is almost always a step towards
shutting the node down. If we allow nodes outside the voting configuration to
be the master then the excluded node will continue to be master, which is
confusing.
This commit adjusts the logic to allow master-eligible nodes to attempt an
election even if they are not in the voting configuration. If such a master is
successfully elected then it adds itself to the voting configuration. This
commit also adjusts the logic that causes master nodes to abdicate when they
are excluded from the voting configuration, to avoid the confusion described
above.
Relates #37712, #37802.
With this change, we will rebuild the live version map and local
checkpoint using documents (including soft-deleted) from the safe commit
when opening an internal engine. This allows us to safely prune away _id
of all soft-deleted documents as the version map is always in-sync with
the Lucene index.
Relates #40741
Supersedes #42979
This commit modifies InternalTestCluster to allow using client() and
other operations inside a RestartCallback (onStoppedNode typically).
Restarting nodes are now removed from the map and thus all
methods now return the state as if the restarting node does not exist.
This avoids various exceptions stemming from accessing the stopped
node(s).
Adds `equals()` and `hashcode()` methods to `DiscoveryNodeRole` to compare
these objects' values for equality, and adds a field to allow us to distinguish
unknown roles from known ones with the same name and abbreviation, for clearer
test failures.
Relates #43175
With this change we only have to add one line to add a new version.
The intent is to make it less error prone and easier to write a script
to automate the process.
There are a few tests within NodeTests that submit items to the
threadpool and then close the node. The tests are designed to check
how running tasks are affected during node close. These tests can cause
CI failures since the submitted tasks may not be running when the node
is closed and then execute after the thread context is closed, which
triggers an unexpected exception. This change ensures the threads are
running so we avoid the unexpected exception and can test these cases.
The test of task submittal while a node is closing is also important so
an additional but muted test has been added that tests the case where a
task may be getting submitted while the node is closing and ensuring we
do not trigger anything unexpected in these cases.
Relates #42774
Relates #42577
This PR proposes to model big integers as longs (and big decimals as doubles)
in the context of dynamic mappings.
Previously, the dynamic mapping logic did not recognize big integers or
decimals, and would an error of the form "No matching token for number_type
[BIG_INTEGER]" when a dynamic big integer was encountered. It now accepts these
numeric types and interprets them as 'long' and 'double' respectively. This
allows `dynamic_templates` to accept and and remap them as another type such as
`keyword` or `scaled_float`.
Addresses #37846.
The test fails if querying the roles via a transport client, since the
transport client does not have the plugin necessary to interpret the additional
role correctly. This commit adds this plugin to the transport client used.
Relates #43175Fixes#43223
Currently the randomization of the q.b. in these tests can create query strings
that can cause caching to be disabled for this query if we query all fields and
there is a date field present. This is pretty much an anomaly that we shouldn't
generally test for in the "testToQuery" tests where cache policies are checked.
This change makes sure we don't create offending query strings so the cache
checks never hit these cases and adds a special test method to check this edge
case.
Closes#43112
roundUp parsers were losing the composite pattern information when new
JavaDateFormatter was created from methods withLocale or withZone.
The roundUp parser should be preserved when calling these methods. This is the same approach in withLocale/Zone methods as in daa2ec8a60/server/src/main/java/org/elasticsearch/common/time/JavaDateFormatter.javacloses#42835
Given the significant performance impact that NIOFS has when term dicts are
loaded off-heap this change enforces FstLoadMode#AUTO that loads term dicts
off heap only if the underlying index input indicates a memory map.
Relates to #43150
As the ValuesSourceType evolves, it is important to be
confident that new enum constants do not break
backwards-compatibility on the stream. Having dedicated
unit tests for this class will help be sure of that.
This change adds the terms index (`.tip`) to the list of extensions
that are memory-mapped by hybridfs. These files used to be accessed
only once to load the terms index on-heap but since #42838 they can
now be used to read the binary FST directly so it is benefical to
memory-map them instead of accessing them via NIO.
Fixes an issue where tests would sometimes hang for 5 seconds when restarting a node. The reason
is that the SeedHostsResolver is blockingly waiting on a result for the full 5 seconds when the
corresponding threadpool is shut down.
We are still using `FileSwitchDirectory` in the case a user configures file based pre-load of mmaps. This is trappy for multiple reasons if the both directories used by `FileSwitchDirectory` point to the same filesystem directory. One issue is LUCENE-8835 that cause issues like #37111 - unless LUCENE-8835 isn't fixed we should not use it in elasticsearch. Instead we use a similar trick as we use for HybridFS and subclass mmap directory directly.
When set to false, allowPartialSearchResults option does not check if the
shard failures have been reseted to null. The atomic array, that is used to record
shard failures, is filled with a null value if a successful request on a shard happens
after a failure on a shard of another replica. In this case the atomic array is not empty
but contains only null values so this shouldn't be considered as a failure since all
shards are successful (some replicas have failed but the retries on another replica succeeded).
This change fixes this bug by checking the content of the atomic array and fails the request only
if allowPartialSearchResults is set to false and at least one shard failure is not null.
Closes#40743
Currently suggesters return null values on empty shards. Usually this gets replaced
by results from other non-epmty shards, but if the index is completely epmty (e.g. after
creation) the search responses "suggest" is also "null" and we don't render a corresponding
output in the REST response. This is an irritating edge case that requires special handling on
the user side (see #42473) and should be fixed.
This change makes sure every suggester type (completion, terms, phrase) returns at least an
empty skeleton suggestion output, even for empty shards. This way, even if we don't find
any suggestions anywhere, we still return and output the empty suggestion.
Closes#42473
When a search on some indices takes a long time, it may cause problems to other indices that are being searched as part of the same search request and being written to as well, because their search context needs to stay open for a long time. This is especially a problem when searching against throttled and non-throttled indices as part of the same request. The problem can be generalized though: this may happen whenever read-only indices are searched together with indices that are being written to. Search contexts staying open for a long time is only an issue for indices that are being written to, in practice.
This commit splits the search in two sub-searches: one for read-only indices, and one for ordinary indices. This way the two don't interfere with each other. The split is done only when size is greater than 0, no scroll is provided and query_then_fetch is used as search type. Otherwise, the search executes like before. Note that the returned num_reduce_phases reflect the number of reduction phases that were run. If the search is split in two, there are three reductions: one non-final for each search, and a final one that merges the results of the previous two.
Closes#40900
The SearchRequest#crossClusterSearch method is currently used only as
part of cross cluster search request, when minimizing roundtrips.
It will soon be used also when splitting a search into two: one for
throttled and one for non throttled indices. It will probably be used
for other usecases as well in the future, hence it makes sense to generalize its name to subSearchRequest.
Though not in use in elasticsearch currently, it seems surprising that
ThreadPool.scheduler().scheduleAtFixedRate would hang. A recurring
scheduled task is never completed (except on failure) and we test for
exceptions using RunnableFuture.get(), which hangs for periodic tasks.
Fixed by checking that task is done before calling .get().
Today the master eagerly reroutes the cluster as part of processing node joins.
However, it is not necessary to do this reroute straight away, and it is
sometimes preferable to defer it until later. For instance, when the master
wins its election it processes joins and performs a reroute, but it would be
better to defer the reroute until after the master has become properly
established.
This change defers this reroute into a separate task, and batches multiple such
tasks together.
#40741 introduced a merge policy that can drop the postings for the `_id`
field on soft deleted documents. The TermsSliceQuery assumes that every document
has has an entry in the postings for that field so it doesn't check if the terms
index exists or not. This change fixes this bug by checking if the terms index for
the `_id` field is null and ignore the segment entirely if it's the case. This should
be harmless since segments without an `_id` terms index should only contain soft deleted
documents.
Closes#42996
If linearizability checking fails with OOM (or other exception), we did
not get the serialized history written into the log, making it difficult
to debug in cases where the problem is hard to reproduce. Fixed to
always attempt dumping the serialized history.
Related to #42244
Both TransportAnalyzeAction and CategorizationAnalyzer have logic to build
custom analyzers for index-independent analysis. A lot of this code is duplicated,
and it requires the AnalysisRegistry to expose a number of internal provider
classes, as well as making some assumptions about when analysis components are
constructed.
This commit moves the build logic directly into AnalysisRegistry, reducing the
registry's API surface considerably.
Setting `auto` after the fuzzy operator (e.g. `"query": "foo~auto"`) in the `query_string`
does not take the length of the term into account when computing the distance and always use
a max distance of 1. This change fixes this disrepancy by ensuring that the term is passed when
the fuzziness is computed.
We respect allocation deciders, including the `MaxRetryAllocationDecider`, when
executing reroute commands. If you specify `?retry_failed=true` then the retry
counter is reset, but today this does not happen until after trying to execute
the reroute commands. This means that if an allocation has repeatedly failed,
but you want to take control and assign a shard to a particular node to work
around the repeated failures, you cannot execute the routing command in the
same call to `POST /_cluster/reroute` as the one that resets the failure
counter.
This commit fixes this by resetting the failure counter first, meaning that you
can now explicitly allocate a repeatedly-failed shard like this:
```
POST /_cluster/reroute?retry_failed=true
{
"commands": [
{
"allocate_replica": {
"index": "blahblah",
"shard": 2,
"node": "node-4"
}
}
]
}
```
Fixes#39546
This commit refactors put mapping request validation for reuse. The
concrete case that we are after here is the ability to apply effectively
the same framework to indices aliases requests. This commit refactors
the put mapping request validation framework to allow for that.
This commit fixes a test bug in the request validators random test. In
particular, an assertion was not properly nested in a guard that would
ensure that was at least one failure.
Relates #43000
When applying put mapping validators, we apply all the validators in the
collection. If a failure occurs, we collect that as a top-level
exception, and suppress any additional failures into the top-level
exception. However, if a request passes the validator after a top-level
exception has been collected, we would try to suppress a null exception
into the top-level exception. This is a violation of the
Throwable#addSuppressed API. This commit addresses this, and adds test
to cover the logic of collecting the failures when validating a put
mapping request.
Today we test for translog corruption by incrementing a byte by 1 somewhere in
a file, and verify that this leads to a `TranslogCorruptionException`.
However, we rely on _all_ corruptions leading to this exception in the
`RemoveCorruptedShardDataCommand`: this command fails if a translog file
corruption leads to a different kind of exception, and `EOFException` and
`NegativeArraySizeException` are both possible. This commit strengthens the
translog corruption detection tests by simulating the following:
- a random value is written
- the file is truncated
It also makes sure that we return a `TranslogCorruptionException` in all such
cases.
Fixes#42661
Backport of #42744
This code has not been needed since the removal of tribe nodes, it was
left behind when those were dropped (note that regular transport
permissions are handled through transport profiles, even if they are not
explicitly in use).
Previously, a reindex request had two different size specifications in the body:
* Outer level, determining the maximum documents to process
* Inside the source element, determining the scroll/batch size.
The outer level size has now been renamed to max_docs to
avoid confusion and clarify its semantics, with backwards compatibility and
deprecation warnings for using size.
Similarly, the size parameter has been renamed to max_docs for
update/delete-by-query to keep the 3 interfaces consistent.
Finally, all 3 endpoints now support max_docs in both body and URL.
Relates #24344
Today we assert that the connection thread is blocked by the time the test gets
to the barrier, but in fact this is not a valid assertion. The following
`Thread.sleep()` will cause the test to fail reasonably often.
```diff
diff --git a/server/src/test/java/org/elasticsearch/cluster/NodeConnectionsServiceTests.java b/server/src/test/java/org/elasticsearch/cluster/NodeConnectionsServiceTests.java
index 193cde3180d..0e57211cec4 100644
--- a/server/src/test/java/org/elasticsearch/cluster/NodeConnectionsServiceTests.java
+++ b/server/src/test/java/org/elasticsearch/cluster/NodeConnectionsServiceTests.java
@@ -364,6 +364,7 @@ public class NodeConnectionsServiceTests extends ESTestCase {
final CheckedRunnable<Exception> connectionBlock = nodeConnectionBlocks.get(node);
if (connectionBlock != null) {
try {
+ Thread.sleep(50);
connectionBlock.run();
} catch (Exception e) {
throw new AssertionError(e);
```
This change relaxes the test to allow some time for the connection thread to
hit the barrier.
Fixes#40170
Changed order of listener invocation so that we notify before
registering search context and notify after unregistering same.
This ensures that count up/down like what we do in ShardSearchStats
works. Otherwise, we risk notifying onFreeScrollContext before notifying
onNewScrollContext (same for onFreeContext/onNewContext, but we
currently have no assertions failing in those).
Closes#28053
The test failed because we had only a single document in the index
that got deleted such that some assertions that expected at least
one live doc failed.
Relates to: #40741
Single updates use a different internal code path than updates that are wrapped in a bulk request.
While working on a refactoring to bring both closer together I've noticed that bulk updates were
failing some of the tests that single updates passed. In particular, bulk updates cause
NullPointerExceptions to be thrown and listeners not being properly notified when being rejected
from the thread pool.
This change adds a merge policy that drops all _id postings for documents that
are marked as soft-deleted but retained across merges. This is usually unnecessary
unless soft-deletes are used with a retention policy since otherwise a merge would
remove deleted documents anyway.
Yet, this merge policy prevents extreme cases where a very large number of soft-deleted
documents are retained and are impacting update performance.
Note, using this merge policy will remove all lookup by ID capabilities for soft-deleted documents.
Adds a metadata field to snapshots which can be used to store arbitrary
key-value information. This may be useful for attaching a description of
why a snapshot was taken, tagging snapshots to make categorization
easier, or identifying the source of automatically-created snapshots.
Some clusters might have been already migrated to version 7 without being warned about the joda-java migration changes.
Deprecation api on that version will give them guidance on what patterns need to be changed.
relates. This change is using the same logic like in 6.8 that is: verifying the pattern is from the incompatible set ('y'-Y', 'C', 'Z' etc), not from predifined set, not prefixed with 8. AND was also created in 6.x. Mappings created in 7.x are considered migrated and should not generate warnings
There is no pipeline check (present on 6.8) as it is impossible to verify when the pipeline was created, and therefore to make sure the format is depracated or not
#42010
The full-text query parsers accept field pattern that are expanded using the mapping.
Alias field are also detected during the expansion but they are not deduplicated with the
concrete fields that are found from other patterns (or the same). This change ensures
that we deduplicate the target fields of the full-text query parsers in order to avoid
adding the same clause multiple times. Boolean queries are already able to deduplicate
clauses during rewrite but since we also use DisjunctionMaxQuery it is preferable to detect
these duplicates early on.
This change makes use of the reader attributes added in LUCENE-8671
to ensure that `_id` fields are always on-heap for best update performance
and term dicts are generally off-heap on Read-Only engines.
Closes#38390
This test is failing because recoveries of these empty shards are not
completing in a reasonable time, but the reason for this is still obscure. This
commit adds yet more logging.
Relates #40174, #42424
This commit adds functionality so that aliases that are manipulated on
leader indices are replicated by the shard follow tasks to the follower
indices. Note that we ignore write indices. This is due to the fact that
follower indices do not receive direct writes so the concept is not
useful.
Relates #41815
This commit performs the proper restore of network disruption.
Previously disruptionScheme.stopDisrupting() was called that does not
ensure that connectivity between cluster nodes is restored. The test
was checking that the cluster has green status, but it was not checking
that connectivity between nodes is restored.
Here we switch to internalCluster().clearDisruptionScheme(true) which
performs both checks before returning.
Closes#39688
(cherry picked from commit c8988d5cf5a85f9b28ce148dbf100aaa6682a757)
The control flow in TransportAnalyzeAction is currently spread across two large
methods, and is quite difficult to follow. This commit tidies things up a bit, to make
it clearer when we use pre-defined analyzers and when we use custom built ones.
This commit clones the existing AnalyzeRequest/AnalyzeResponse classes
to the high-level rest client, and adjusts request converters to use these new
classes.
This is a prerequisite to removing the Streamable interface from the internal
server version of these classes.
IntervalBuilder#analyzeText will currently return null if it is passed an
empty TokenStream, which can lead to a confusing NullPointerException
later on during querying. This commit changes the code to return
NO_INTERVALS instead.
Fixes#42587
We were checking if an exception was caused by a specific reason "Not a
directory". Alas, this reason is locale-dependent and can fail on
systems that are not set to en_US.UTF-8. This commit addresses this by
deriving what the locale-dependent error message would be and using that
for comparison with the actual exception thrown.
Relates #41689
We had this as a dependency for legacy dependencies that still needed
the Log4j 1.2 API. This appears to no longer be necessary, so this
commit removes this artifact as a dependency.
To remove this dependency, we had to fix a few places where we were
accidentally relying on Log4j 1.2 instead of Log4j 2 (easy to do, since
both APIs were on the compile-time classpath).
Finally, we can remove our custom Netty logger factory. This was needed
when we were on Log4j 1.2 and handled logging in our own unique
way. When we migrated to Log4j 2 we could have dropped this
dependency. However, even then Netty would still pick up Log4j 1.2 since
it was on the classpath, thus the advantage to removing this as a
dependency now.
Today Elasticsearch does not prevent you from reconfiguring a node's
`path.data` to point to data paths that previously belonged to more than one
node. There's no good reason to be able to do this, and the consequences can be
quietly disastrous. Furthermore, #42489 might result in a user trying to split
up a previously-shared collection of data paths by hand and there's definitely
scope for mixing the paths up across nodes when doing this.
This change adds a check during startup to ensure that each data path belongs
to the same node.
Since the max_score optimization landed in Elasticsearch 7,
the CommonTermsQuery is redundant and slower. Moreover the
cutoff_frequency parameter for MatchQuery and MultiMatchQuery
is redundant.
Relates to #27096
(cherry picked from commit 04b74497314eeec076753a33b3b6cc11549646e8)
Today the `LeaderChecker` and `HandshakingTransportAddressConnector` do not log
anything above `DEBUG` level. However there are some situations where it is
appropriate for them to log at a higher level:
- if the low-level handshake succeeds but the high-level one fails then this
indicates a config error that the user should resolve, and the exception
will help them to do so.
- if leader checks fail repeatedly then we restart discovery, and the exception
will help to determine what went wrong.
Resolves#42153
Refactors the WKT and GeoJSON parsers from an utility class into an
instantiatable objects. This is a preliminary step in
preparation for moving out coordinate validators from Geometry
constructors. This should allow us to make validators plugable.
A disruption test case need to use a lower checkpoint sync interval
since they verify sequence numbers after the test waiting max 10 seconds
for it to stabilize.
Closes#42637
* Now that we process the bulk requests themselves on the WRITE threadpool, they can run out of retries too like the item requests even when backoff is active
* Fixes#41324 by using the same logic that checks failed item requests for their retry status for the top level bulk requests as well
When multiple commands are called in sequence, fetch shards
from mutable, up-to-date routing nodes to ensure each command's
changes are visible to subsequent commands.
This addresses an issue uncovered during work on #41050.
The problem this commit addresses is that state recovery is not reset on a node that then becomes
master with a cluster state that has a state not recovered flag in it. The situation that was observed
in a failed test run of MinimumMasterNodesIT.testThreeNodesNoMasterBlock (see below) is that we
have 3 master nodes (node_t0, node_t1, node_t2), two of them are shut down (node_t2 remains),
when the first one comes back (renamed to node_t4) it becomes leader in term 2 and sends state
(with state_not_recovered_block) to node_t2, which accepts. node_t2 becomes leader in term 3, and
as it was previously leader in term1 and successfully completed state recovery, does never retry
state recovery in term 3.
Closes#39172
* Make unwrapCorrupt Check Suppressed Ex. (#41889)
* As discussed in #24800 we want to check for suppressed corruption
indicating exceptions here as well to more reliably categorize
corruption related exceptions
* Closes#24800, 41201
* Cleanup Bulk Delete Exception Logging
* Follow up to #41368
* Collect all failed blob deletes and add them to the exception message
* Remove logging of blob name list from caller exception logging
If all primary shards are allocated on the master node, then the
verifying before close step will never interact with mock transport
service. This change prefers to allocate shards on data-only nodes.
Closes#39757
* It looks like we might be cancelling a previous publication instead of
the one triggered by the given request with a very low likelihood.
* Fixed by adding a wait for no in-progress publications
* Also added debug logging that would've identified this problem
* Closes#36813
* Remove Delete Method from BlobStore (#41619)
* The delete method on the blob store was used almost nowhere and just duplicates the delete method on the blob containers
* The fact that it provided for some recursive delete logic (that did not behave the same way on all implementations) was not used and not properly tested either
* Added separate enum for the state of each shard, it was really
confusing that we used the same enum for the state of the snapshot
overall and the state of each individual shard
* relates https://github.com/elastic/elasticsearch/pull/40943#issuecomment-488664150
* Shortened some obvious spots in equals method and saved a few lines
via `computeIfAbsent` to make up for adding 50 new lines to this class
* Remove Obsolete BwC Logic from BlobStoreRepository
* We can't restore 1.3.3 files anyway -> no point in doing the dance of computing a hash here
* Some other minor+obvious cleanups
* Some Cleanup in o.e.i.engine
* Remove dead code and parameters
* Reduce visibility in some obvious spots
* Add missing `assert`s (not that important here since the methods
themselves will probably be dead-code eliminated) but still
If a single pipeline is updated then the internal representation of
all pipelines was updated. With this change, only the internal representation
of the pipelines that have been modified will be updated.
Prior to this change the IngestMetadata of the previous and current cluster
was used to determine whether the internal representation of pipelines
should be updated. If applying the previous cluster state change failed then
subsequent cluster state changes that have no changes to IngestMetadata
will not attempt to update the internal representation of the pipelines.
This commit, changes how the IngestService updates the internal representation
by keeping track of the underlying configuration and use that to detect
against the new IngestMetadata whether a pipeline configuration has been
changed and if so, then the internal pipeline representation will be updated.
We need more information to understand why CcrRetentionLeaseIT is
failing. This commit adds some debug log to retention leases and enables
them in CcrRetentionLeaseIT.
This commit removes the act of renewing some retention leases during a
retention lease recovery test. Having renewal does not add anything
extra to this test, but does allow for some situations where the test
can fail spuriously (i.e., in a way that does not indicate that
production code is broken).
If we close an engine while a refresh is happening, then we might leak
refCount of some SegmentReaders. We need to skip the ram accounting
circuit breaker check until we have a new Lucene snapshot which includes
the fix for LUCENE-8809.
This also adds a test to the engine but left it muted so we won't forget
to reenable this check.
Closes#30290
The shard CLI tool would not do anything if a corruption marker was not
present. But a corruption marker is only added if a corruption is
detected during indexing/writing, not if a search or other read fails.
Changed the tool to always check shards regardless of corruption marker
presence.
Related to #41298
Previously sorting on a missing nested field would fail with an
Exception:
`[nested_field] failed to find nested object under path [nested_path]`
despite `unmapped_type` being set on the query.
Fixes: #33644
(cherry picked from commit 631142d5dd088a10de8dcd939b50a14301173283)
Today the default stabilisation time is calculated on the assumption that the
elected master has no pending tasks to process when it is elected, but this is
not a safe assumption to make. This can result in a cluster reaching the end of
its stabilisation time without having stabilised. Furthermore in #36943 we
increased the probability that each step in `runRandomly()` enqueues another
task, vastly increasing the chance that we hit such a situation.
This change extends the stabilisation process to allow time for all pending
tasks, plus a task that might currently be in flight.
Fixes#41967, in which the master entered the stabilisation phase with over 800
tasks to process.
By default, we track total hits up to 10k but we might index more than
10k documents `testPrimaryRelocationWhileIndexing`. With this change, we
always request for the accurate total hits in the test.
> java.lang.AssertionError: Count is 10000+ hits but 11684 was expected.
Both of these classes are basically a bloated wrapper around a simple
construct that can simply be a DirectoryFactory interface. This change
removes both classes and replaces them with a simple stateless interface
that creates a new `Directory` per shard. The concept of `index.store` is preserved
since it makes sense from a configuration perspective.
Today the `TransportClusterStateAction` ignores the state passed by the
`TransportMasterNodeAction` and obtains its state from the cluster applier.
This might be inconsistent, showing a different node as the master or maybe
even having no master.
This change adjusts the action to use the passed-in state directly, and adds
tests showing that the state returned is consistent with our expectations even
if there is a concurrent master failover.
Fixes#38331
Relates #38432
Today `RetentionLeaseIT` calls `fail(e.toString())` on some exceptions, losing
the stack trace that came with the exception. This commit adjusts this to
re-throw the exception wrapped in an `AssertionError` so we can see more
details about failures such as #41430.
This commit adds a log message containing the routing table, emitted on each
iteration of the failing assertBusy() in #40174. It also modernizes the code a
bit.
This commit makes creators of GetField split the fields into document fields and metadata fields. It is part of larger refactoring that aims to remove the calls to static methods of MapperService related to metadata fields, as discussed in #24422.
`org.elasticsearch.action.bulk.BulkProcessor` is a threadsafe class that
allows for simple semantics to deal with sending bulk requests. Once a
bulk reaches it's pre-defined size, documents, or flush interval it will
execute sending the bulk. One configurable option is the number of concurrent
outstanding bulk requests. That concurrency is implemented in
`org.elasticsearch.action.bulk.BulkRequestHandler` via a semaphore. However,
the only code that currently calls into this code is blocked by `synchronized`
methods. This results in the in-ability for the BulkProcessor to behave concurrently
despite supporting configurable amounts of concurrent requests.
This change removes the `synchronized` method in favor an explicit
lock around the non-thread safe parts of the method. The call into
`org.elasticsearch.action.bulk.BulkRequestHandler` is no longer blocking, which
allows `org.elasticsearch.action.bulk.BulkRequestHandler` to handle it's own concurrency.
Lucene 8 has the ability to skip blocks of non-competitive documents.
However some queries don't track their maximum score (`script_score`, `span`, ...)
so they always return Float.POSITIVE_INFINITY as maximum score. This can slow down
some boolean queries if other clauses have bounded max scores. This commit disables
the max score optimization when we detect a mandatory scoring clause with unbounded
max scores. Optional clauses are not checked since they can still skip documents
when the unbounded clause is after the current document.
Currently AnalysisRegistry#processNormalizerFactory creates a normalizer and
only later checks whether it should be added to the normalizer map passed in. In
case we throw an exception it isn't closed. This can be prevented by moving the
check that throws the exception earlier.
Allow for SimpleQueryString, QueryString and MultiMatchQuery
to set the `fields` parameter to the wildcard `*`. If so, set
the leniency to `true`, to achieve the same behaviour as from the
`"default_field" : "*" setting.
Furthermore, check if `*` is in the list of the `default_field` but
not necessarily as the 1st element.
Closes: #39577
(cherry picked from commit e75ff0c748e6b68232c2b08e19ac4a4934918264)
AbstractDisruptionTestCase set a lower global checkpoint sync interval setting, but this was ignored by
testAckedIndexing, which has led to spurious test failures
Relates #41068, #38931
ShardId already implements Writeable so there is no need for it to implement Streamable too. Also the readShardId static method can be
easily replaced with direct usages of the constructor that takes a
StreamInput as argument.
In case a search request holds only the suggest section, the query phase
is skipped and only the suggest phase is executed instead. There will
never be hits returned, and in case the explain flag is set to true, the
explain sub phase throws a null pointer exception as the query is null.
Usually a null query is replaced with a match all query as part of SearchContext#preProcess which is though skipped as well with suggest
only searches. To address this, we skip the explain sub fetch phase
for search requests that only requested suggestions.
Closes#31260
Removing of payload in BulkRequest (#39843) had a side effect of making
`BulkRequest.add(DocWriteRequest<?>...)` (with varargs) recursive, thus
leading to StackOverflowError. This PR adds a small change in
RequestConvertersTests to show the error and the corresponding fix in
`BulkRequest`.
Fixes#41668
* Remove IndexShard dependency from Repository
In order to simplify repository testing especially for BlobStoreRepository
it's important to remove the dependency on IndexShard and reduce it to
Store and MapperService (in the snapshot case). This significantly reduces
the dependcy footprint for Repository and allows unittesting without starting
nodes or instantiate entire shard instances. This change deprecates the old
method signatures and adds a unittest for FileRepository to show the advantage
of this change.
In addition, the unittesting surfaced a bug where the internal file names that
are private to the repository were used in the recovery stats instead of the
target file names which makes it impossible to relate to the actual lucene files
in the recovery stats.
* don't delegate deprecated methods
* apply comments
* test
When using High Level Rest Client Java API to produce search query, using AggregationBuilders.topHits("th").sort("_score", SortOrder.DESC)
caused query to contain duplicate sort clauses.
A shard that is undergoing peer recovery is subject to logging warnings of the form
org.elasticsearch.action.FailedNodeException: Failed node [XYZ]
...
Caused by: org.apache.lucene.index.IndexNotFoundException: no segments* file found in ...
These failures are actually harmless, and expected to happen while a peer recovery is ongoing (i.e.
there is an IndexShard instance, but no proper IndexCommit just yet).
As these failures are currently bubbled up to the master, they cause unnecessary reroutes and
confusion amongst users due to being logged as warnings.
Closes #40107
Moves the test infrastructure away from using node.max_local_storage_nodes, allowing us in a
follow-up PR to deprecate this setting in 7.x and to remove it in 8.0.
This also changes the behavior of InternalTestCluster so that starting up nodes will not automatically
reuse data folders of previously stopped nodes. If this behavior is desired, it needs to be explicitly
done by passing the data path from the stopped node to the new node that is started.
Simplifies the voting configuration reconfiguration logic by switching to an explicit Comparator for
the priorities. Does not make changes to the behavior of the component.
Flushing at the end of a peer recovery (if needed) can bring these
benefits:
1. Closing an index won't end up with the red state for a recovering
replica should always be ready for closing whether it performs the
verifying-before-close step or not.
2. Good opportunities to compact store (i.e., flushing and merging
Lucene, and trimming translog)
Closes#40024Closes#39588
This change verifies and aborts recovery if source and target have the
same syncId but different sequenceId. This commit also adds an upgrade
test to ensure that we always utilize syncId.
The verifying-before-close step ensures the global checkpoints on all
shard copies are in sync; thus, we don' t need to sync global
checkpoints for closed indices.
Relate #33888
There is an off-by-one error in this test. It leads to the recovery
thread never being started, and that means joining on it will wait
indefinitely. This commit addresses that by fixing the off-by-one error.
Relates #42325
I forgot to git add these before pushing, sorry. This commit fixes
compilation in IndexShardTests, they are needed here and not in master
due to differences in how Java infers types in generics between JDK 8
and JDK 11.
Today when executing an action on a primary shard under permit, we do
not enforce that the shard is in primary mode before executing the
action. This commit addresses this by wrapping actions to be executed
under permit in a check that the shard is in primary mode before
executing the action.
Today we are persisting the retention leases at least every thirty
seconds by a scheduled background sync. This sync causes an fsync to
disk and when there are a large number of shards allocated to slow
disks, these fsyncs can pile up and can severely impact the system. This
commit addresses this by only persisting and fsyncing the retention
leases if they have changed since the last time that we persisted and
fsynced the retention leases.
* Cleanup Various Uses of ActionListener
* Use shorter `map`, `runAfter` or `wrap` where functionally equivalent to anonymous class
* Use ActionRunnable where functionally equivalent
Downgrading an Elasticsearch node to an earlier version is unsupported, because
we do not make any attempt to guarantee that a node can read any of the on-disk
data written by a future version. Yet today we do not actively prevent
downgrades, and sometimes users will attempt to roll back a failed upgrade with
an in-place downgrade and get into an unrecoverable state.
This change adds the current version of the node to the node metadata file, and
checks the version found in this file against the current version at startup.
If the node cannot be sure of its ability to read the on-disk data then it
refuses to start, preserving any on-disk data in its upgraded state.
This change also adds a command-line tool to overwrite the node metadata file
without performing any version checks, to unsafely bypass these checks and
recover the historical and lenient behaviour.
Secure settings currently error if they exist inside elasticsearch.yml.
This commit adds validation that non-secure settings do not exist inside
the keystore.
closes#41831
The date_histogram accepts an interval which can be either a calendar
interval (DST-aware, leap seconds, arbitrary length of months, etc) or
fixed interval (strict multiples of SI units). Unfortunately this is inferred
by first trying to parse as a calendar interval, then falling back to fixed
if that fails.
This leads to confusing arrangement where `1d` == calendar, but
`2d` == fixed. And if you want a day of fixed time, you have to
specify `24h` (e.g. the next smallest unit). This arrangement is very
error-prone for users.
This PR adds `calendar_interval` and `fixed_interval` parameters to any
code that uses intervals (date_histogram, rollup, composite, datafeed, etc).
Calendar only accepts calendar intervals, fixed accepts any combination of
units (meaning `1d` can be used to specify `24h` in fixed time), and both
are mutually exclusive.
The old interval behavior is deprecated and will throw a deprecation warning.
It is also mutually exclusive with the two new parameters. In the future the
old dual-purpose interval will be removed.
The change applies to both REST and java clients.
If `keyedFilters` is null it assumes there are unkeyed filters...which
will NPE if the unkeyed filters was actually empty.
This refactors to simplify the filter assignment a bit, adds an empty
check and tidies up some formatting.
This commit fixes a test bug that ends up comparing the result of two consecutive calls to System.currentTimeMillis that can be different
on slow CIs.
Closes#42064
This test can create and shuffle 2*(3*5*7) = 210 shards which is quite
heavy for our CI. This commit reduces the load, so we don't timeout on
CI.
Closes#28153
Adds an initial limited implementations of geo features to SQL. This implementation is based on the [OpenGIS® Implementation Standard for Geographic information - Simple feature access](http://www.opengeospatial.org/standards/sfs), which is the current standard for GIS system implementation. This effort is concentrate on SQL option AKA ISO 19125-2.
Queries that are supported as a result of this initial implementation
Metadata commands
- `DESCRIBE table` - returns the correct column types `GEOMETRY` for geo shapes and geo points.
- `SHOW FUNCTIONS` - returns a list that includes supported `ST_` functions
- `SYS TYPES` and `SYS COLUMNS` display correct types `GEO_SHAPE` and `GEO_POINT` for geo shapes and geo points accordingly.
Returning geoshapes and geopoints from elasticsearch
- `SELECT geom FROM table` - returns the geoshapes and geo_points as libs/geo objects in JDBC or as WKT strings in console.
- `SELECT ST_AsWKT(geom) FROM table;` and `SELECT ST_AsText(geom) FROM table;`- returns the geoshapes ang geopoints in their WKT representation;
Using geopoints to elasticsearch
- The following functions will be supported for geopoints in queries, sorting and aggregations: `ST_GeomFromText`, `ST_X`, `ST_Y`, `ST_Z`, `ST_GeometryType`, and `ST_Distance`. In most cases when used in queries, sorting and aggregations, these function are translated into script. These functions can be used in the SELECT clause for both geopoints and geoshapes.
- `SELECT * FROM table WHERE ST_Distance(ST_GeomFromText(POINT(1 2), point) < 10;` - returns all records for which `point` is located within 10m from the `POINT(1 2)`. In this case the WHERE clause is translated into a range query.
Limitations:
Geoshapes cannot be used in queries, sorting and aggregations as part of this initial effort. In order to fully take advantage of geoshapes we would need to have access to geoshape doc values, which is coming in #37206. `ST_Z` cannot be used on geopoints in queries, sorting and aggregations since we don't store altitude in geo_point doc values.
Relates to #29872
Backport of #42031
This change updates tests that use a CountDownLatch to synchronize the
running of threads when testing concurrent operations so that we ensure
the thread has been fully created and run by the scheduler. Previously,
these tests used a latch with a value of 1 and the test thread counted
down while the threads performing concurrent operations just waited.
This change updates the value of the latch to be 1 + the number of
threads. Each thread counts down and then waits. This means that each
thread has been constructed and has started running. All threads will
have a common start point now.
Today we do not expose the cluster UUID in any logs by default, but it would be
useful to see it. For instance if a user starts multiple nodes as separate
clusters then they will silently remain as separate clusters even if they are
subsequently reconfigured to look like a single cluster. This change logs the
committed cluster UUID the first time the node encounters it.
* Switch to using a list instead of a Set for the filters, so that the
order of these filters is kept.
(cherry picked from commit 74a743829799b64971e0ac5ae265f43f6c14e074)
If remote recovery copies an index commit which has gaps in sequence
numbers to a follower; then these assertions (introduced in #40823)
don't hold for follower replicas.
Closes#41037
Currently IndexAnalyzers keeps the three default as separate class members
although they should refer to the same analyzers held in the additional
analyzers map under the default names. This assumption should be made more
explicit by keeping all analyzers in the map. This change adapts the constructor
to check all the default entries are there and the getters to reach into the map
with the default names when needed.
The changes in #39317 brought to light some concurrency issues in the
close method of Recyclers as we do not wait for threads running in the
threadpool to be finished prior to the closing of the PageCacheRecycler
and the Recyclers that are used internally. #41695 was opened to
address the concurrent close issues but upon review, the closing of
these classes is not really needed as the instances should be become
available for garbage collection once there is no longer a reference to
the closed node.
Closes#41683
We have a number of places in analysis-handling code where we check
if a field type is a keyword field, and if so then extract the normalizer rather
than pulling the index-time analyzer. However, a keyword normalizer is
really just a special case of an analyzer, so we should be able to simplify this
by setting the normalizer as the index-time analyzer at construction time.
* This test was resulting in a `PARTIAL` instead of a `SUCCESS` state for
the case of closing an index during snapshotting on 7.x
* The reason for this is the changed default behaviour regarding
waiting for active shards between 8.0 and 7.x
* Fixed by adjusting the waiting behaviour on the close index request
in the test
* Closes#39828
Yannick Welsch
Tanguy Leroux
Luca Cavanna
Jim Ferenczi
Jason Tedor
Lee Hinman
Christos Soulios
Armin Braun
Nhat Nguyen
Martijn van Groningen
Paul Sanwald
Igor Motov
David Turner
Henning Andersen
Alpar Torok
Jay Modi
Julie Tibshirani
Christoph Büscher
Przemyslaw Gomulka
Przemysław Witek
Simon Willnauer
Tal Levy
Alan Woodward
Alexander Reelsen
Jim Ferenczi
Vigya Sharma
markharwood
henryptung
Gordon Brown
Mark Vieira
Andrey Ershov
Yu
Marios Trivyzas
Adrien Grand
Vigya Sharma
Daniel Mitterdorfer
weizijun
Jack Conradson
Ryan Ernst
Jake Landis
Guillaume Darmont
mushao999
Ignacio Vera
Zachary Tong
Yogesh Gaikwad
Andrei Stefan