This removes Elasticsearch's filter cache and uses Lucene's instead. It has some
implications:
- custom cache keys (`_cache_key`) are unsupported
- decisions are made internally and can't be overridden by users ('_cache`)
- not only filters can be cached but also all queries that do not need scores
- parent/child queries can now be cached, however cached entries are only
valid for the current top-level reader so in practice it will likely only
be used on read-only indices
- the cache deduplicates filters, which plays nicer with large keys (eg. `terms`)
- better stats: we already had ram usage and evictions, but now also hit count,
miss count, lookup count, number of cached doc id sets and current number of
doc id sets in the cache
- dynamically changing the filter cache size is not supported anymore
Internally, an important change is that it removes the NoCacheFilter infrastructure
in favour of making Query.rewrite specializing the query for the current reader so
that it will only be cached on this reader (look for IndexCacheableQuery).
Note that consuming filters with the query API (createWeight/scorer) instead of
the filter API (getDocIdSet) is important for parent/child queries because
otherwise a QueryWrapperFilter(ParentQuery) would run the wrapped query per
segment while relations might be cross segments.
Remove the ability to specify search type ‘query_and_fetch’ and
‘df_query_and_fetch’ from the REST API.
- Adds REST tests
- Updates REST API spec to remove ‘query_and_fetch’ and
‘df_query_and_fetch’ as options
- Removes documentation for these options
Closes#9606
The current implementation is dangerous: it unexpectedly refreshes,
which can quickly cause an unhealthy index (segment explosion). It
can also delete different documents on primary vs replicas, causing
inconsistent replicas.
For 2.0 we will replace this with an optional plugin that does a
scan/scroll search and then issues bulk delete requests.
Closes#10859
This commit adds support for structural errors / failures / exceptions
on the elasticsearch REST layer. Exceptions are rendering with at least
a `type` and a `reason` corresponding to the exception name and the message.
Some expcetions like the ones associated with an index or a shard will have
additional information about the index the exception was triggered on or the
shard respectivly.
Each rendered response will also contain a list of root causes which is a list
of distinct shard level errors returned for the request. Root causes are the lowest
level elasticsearch exception found per shard response and are intended to be displayed
to the user to indicate the soruce of the exception.
Shard level response are by-default grouped by their type and reason to reduce the amount
of duplicates retunred. Yet, the same exception retunred from different indices will not be
grouped.
Closes#3303
This commit splits the current ClusterBlockLevel.METADATA into two disctins ClusterBlockLevel.METADATA_READ and ClusterBlockLevel.METADATA_WRITE blocks. It allows to make a distinction between
an operation that modifies the index or cluster metadata and an operation that does not change any metadata.
Before this commit, many operations where blocked when the cluster was read-only: Cluster Stats, Get Mappings, Get Snapshot, Get Index Settings, etc. Now those operations are allowed even when
the cluster or the index is read-only.
Related to #8102, #2833Closes#3703Closes#5855Closes#10521Closes#10522
The field stats api returns field level statistics such as lowest, highest values and number of documents that have at least one value for a field.
An api like this can be useful to explore a data set you don't know much about. For example you can figure at with the lowest and highest response times are, so that you can create a histogram or range aggregation with sane settings.
This api doesn't run a search to figure this statistics out, but rather use the Lucene index look these statics up (using Terms class in Lucene). So finding out these stats for fields is cheap and quick.
The min/max values are based on the type of the field. So for a numeric field min/max are numbers and date field the min/max date and other fields the min/max are term based.
Closes#10523
Also changed the stash logger to not log all stashed values under debug (it does trace now) but do dump the stash content upon failure (under info as a XContent)
Extends ShardStats with commit specific information. We currently expose commit id, generation and the user data map.
The information is also retrievable via the Rest API by using `GET _stats?level=shards`
Closes#10687
In Lucene 5.1 lots of filters got deprecated in favour of equivalent queries.
Additionally, random-access to filters is now replaced with approximations on
scorers. This commit
- replaces the deprecated NumericRangeFilter, PrefixFilter, TermFilter and
TermsFilter with NumericRangeQuery, PrefixQuery, TermQuery and TermsQuery,
wrapped in a QueryWrapperFilter
- replaces XBooleanFilter, AndFilter and OrFilter with a BooleanQuery in a
QueryWrapperFilter
- removes DocIdSets.isBroken: the new two-phase iteration API will now help
execute slow filters efficiently
- replaces FilterCachingPolicy with QueryCachingPolicy
Close#8960
This commit changes dynamic mappings updates so that they are synchronous on the
entire cluster and their validity is checked by the master node. There are some
important consequences of this commit:
- a failing index request on a non-existing type does not implicitely create
the type anymore
- dynamic mappings updates cannot create inconsistent mappings on different
shards
- indexing requests that introduce new fields might induce latency spikes
because of the overhead to update the mappings on the master node
Close#8688
This option defaults to false, because it is also important to upgrade
the "merely old" segments since many Lucene improvements happen within
minor releases.
But you can pass true to do the minimal work necessary to upgrade to
the next major Elasticsearch release.
The HTTP GET upgrade request now also breaks out how many bytes of
ancient segments need upgrading.
Closes#10213Closes#10540
Conflicts:
dev-tools/create_bwc_index.py
rest-api-spec/api/indices.upgrade.json
src/main/java/org/elasticsearch/action/admin/indices/optimize/OptimizeRequest.java
src/main/java/org/elasticsearch/action/admin/indices/optimize/ShardOptimizeRequest.java
src/main/java/org/elasticsearch/action/admin/indices/optimize/TransportOptimizeAction.java
src/main/java/org/elasticsearch/index/engine/InternalEngine.java
src/test/java/org/elasticsearch/bwcompat/StaticIndexBackwardCompatibilityTest.java
src/test/java/org/elasticsearch/index/engine/InternalEngineTests.java
src/test/java/org/elasticsearch/rest/action/admin/indices/upgrade/UpgradeReallyOldIndexTest.java
For bacwards compatibility reasons routing_nodes were previously printed out when routing_table was requested, together with the actual routing_table. Now they are printed out only when requests through `routing_nodes` flag.
Relates to #10412Closes#10486
Cluster state api returns both routing_table and routing_nodes sections whenever routing_table is requested. That is pretty much the same info, just grouped differently. This commit allows to differentiate between the two. Yet, routing_table still returns both for bw comp reasons.
Closes#10352Closes#10412
Align get indexed scripts and get search template apis to our get api, which returns a response body when the document is not found, with a found boolean flag. Also, return metadata info all the time too.
Closes#7325Closes#10396
RoutingTables activePrimaryShardsGrouped(), allActiveShardsGrouped() and
allAssignedShardsGrouped() methods treated empty index array input
parameters as meaning "all" indices and expanded to the routing maps
keyset. However, the expansion of index names is now already done in
MetaData#concreteIndices(). Returning an empty index name list here
when a wildcard pattern didn't match any index name could lead to
problems like #9081 because the RoutingTable still expanded this
list of names to "_all". In case of e.g. the recovery endpoint this
could lead to problems.
Closes#9081Closes#10148
This commit brings the benefits of the `count` search type to search requests
that have a `size` of 0:
- a single round-trip to shards (no fetch phase)
- ability to use the query cache
Since `count` now provides no benefits over `query_then_fetch`, it has been
deprecated.
Close#7630
Deleting a type from an index is inherently dangerous because
the type can be recreated with new mappings which may conflict
with existing segments still using the old mappings. This
removes the ability to delete a type (similar to how deleting
fields within a type is not allowed, for the same reason).
closes#8877closes#10231
This commit adds the current total number of translog operations to the recovery reporting API. We also expose the recovered / total percentage:
```
"translog": {
"recovered": 536,
"total": 986,
"percent": "54.3%",
"total_time": "2ms",
"total_time_in_millis": 2
},
```
Closes#9368Closes#10042
The analysis chain should be used instead of relying on this, as it is
confusing when dealing with different per-field analysers.
The `locale` option was only used for `lowercase_expanded_terms`, which,
once removed, is no longer needed, so it was removed as well.
Fixes#9978
Relates to #9973
We've been relying on URI for url encoding, but it turns out it has some problems. For instance '+' stays as is while it should be encoded to `%2B`. If we go and manually encode query params we have to be careful though not to run into double encoding ('+'=>'%2B'=>'%252B'). The applied solution relies on URI encoding for the url path, but manual url encoding for the query parameters. We prevent URI from double encoding query params by using its single argument constructor that leaves everything as is.
We can also revert back the expression script REST test that revealed this to its original content (which contains an addition).
Closes#9769Closes#9946
While the parser allowed changing field type settings, these would never
have been serialized. So this change simply removes parsing using
parseField. Backcompat will still work if a user uploads old settings
(they just would never have worked anyways, so we continue ignoring
them with 1.x, and 2.x will now error).
see #8143closes#9914
To support the `_recovery` API, the recovery process keeps track of current progress in a class called RecoveryState. This class currently have some issues, mostly around concurrency (see #6644 ). This PR cleans it up as well as other issues around it:
- Make the Index subsection API cleaner:
- remove redundant information - all calculation is done based on the underlying file map
- clearer definition of what is what: total files, vs reused files (local files that match the source) vs recovered files (copied over). % based progress is reported based on recovered files only.
- cleaned up json response to match other API (sadly this breaks the structure). We now properly report human values for dates and other units.
- Add more robust unit testing
- Detail flag was passed along as state (it's now a ToXContent param)
- State lookup during reporting is now always done via the IndexShard , no more fall backs to many other classes.
- Cleanup APIs around time and move the little computations to the state class as opposed to doing them out of the API
I also improved error messages out of the REST testing infra for things I run into.
Closes#6644Closes#9811
The number of current pending tasks is useful to detect and overloaded master. This commit adds it to the cluster health API. The complete list can be retrieved from the dedicated pending tasks API.
It also adds rest tests for the cluster health variants.
Closes#9877
The plus sign is not treated correctly in encoding and can lead
to problems, if the search request is encoded as HTTP parameter
instead of the HTTP body.
Relates #9769
Using '_cat/segments' or the indices segments api without matching any index
now returns empty result instead of throwing IndexMissingException.
Closes#9219
Whenever we have an api that supports GET with a body, we always support the POST method too, as well as providing the body as a query_string parameter called `source`. Our REST spec should reflect this convention. FIxed them and introduced a hard check at parse time in our Java REST tests runner, which will cause the tests to fail if spec are not compliant.
Closes#9629
The percolate api doesn't parse the encoded body provided as `source` query string parameter, when percolating an existing document. Fixed and added REST test that would have caught this since we randomly use GET + encoded `source` param instead of GET + request body in our java runner (the perl runner does the same too).
Closes#9628
The `full` option and `FlushType.NEW_WRITER` only exists to allow
realtime changes to two settings (`index.codec` and `index.concurrency`).
Those settings are very expert and don't really need to be updateable
in realtime.
Until now, there was no possibility to expose infos about configured
transport profiles. This commit adds the ability to expose those
information in the TransportInfo class.
The channel was well as the netty pipeline handler now also contain
the profile they were configured for, as this information cannot be
extracted elsewhere.
In addition, each profile now can set its own publish host and port,
which might be needed in case of portforwarding or using docker.
Closes#9134
Adding missing support for the multi-index query parameters 'ignore_unavailable',
'allow_no_indices' and 'expand_wildcards' to '_cluster/state' API. These
parameters are supposed to be supported for APIs that work across multiple indices.
So far overwriting the default settings per REST call was not possible which is
fixed here.
Closes#5229Closes#9295
Today we give the HTTP status back within the HTTP response itself and within the JSON response as well:
```sh
curl localhost:9200/
```
```js
{
"status" : 200,
"name" : "Red Wolf",
"version" : {
"number" : "2.0.0",
"build_hash" : "6837a61d8a646a2ac7dc8da1ab3c4ab85d60882d",
"build_timestamp" : "2014-08-19T13:55:56Z",
"build_snapshot" : true,
"lucene_version" : "4.9"
},
"tagline" : "You Know, for Search"
}
```
The header indicates to how many shard copies (primary and replicas shards) a write was supposed to go to, to how many
shard copies to write succeeded and potentially captures shard failures if writing into a replica shard fails.
For async writes it also includes the number of shards a write is still pending.
Closes#7994
This fix ensures that calls to the GET alias/mappings/settings/warmers APIs return the aliases/mappings/settings/warmers object even if there is no content within them.. This make them consistent with the GET Index API docs and the breaking changes in 1.4 docs
Closes#9148
Add a new ignore_idle_threads boolean option (default true) to
/_nodes/hot_threads, to filter out threads in known idle places like
waiting on a socket select or on pulling the next task from an empty
queue.
Closes#8985Closes#8908
This commit adds support for version and version_type to the Term Vectors API.
This could be useful in the following case whereby the user gets a document
and later wants to generate its TVs. With version, this would ensure that only
the TVs of that particular document are generated, and error out if the
document has been updated in between.
Closes#7480
Adds a `ignore_like` parameter to the MLT Query, which simply tells the
algorithm to skip all the terms from the given documents. This could be useful
in order to better guide nearest neighbor search by telling the algorithm to
never explore the space spanned by the given `ignore_like` docs. In essence we
are interested about the characteristic of a given item, but not of the ones
provided by `ignore_like`, thereby forcing the algorithm to go deeper in its
selection of terms. Note that this is different than simply performing a must
not boolean query on the unliked items. The syntax is exactly the same as the
`like` parameter.
Closes#8674
Today, Elasticsearch has a separate merge thread pool checking once
per second (by default) if any merges are necessary, but this is no
longer necessary since we can and do now tell Lucene's
ConcurrentMergeScheduler never to "hard pause" threads when merges
fall behind, since we do our own index throttling.
This change goes back to letting Lucene launch merges as needed, and
removes these two expert settings:
index.merge.force_async_merge
index.merge.async_interval
Now merges kick off immediately instead of waiting up to 1 second
before running.
Closes#8643
We speak of the term vectors of a document, where each field has an associated
stored term vector. Since by default we are requesting all the term vectors of
a document, the HTTP request endpoint should rather be called `_termvectors`
instead of `_termvector`. The usage of `_termvector` is now deprecated, as
well as the transport client call to termVector and prepareTermVector.
Closes#8484
Fixed behaviour where two representations of the default index analyzer weren't being treated as equivalent. Added REST test to confirm fix.
Closes#2716
If a shard (e.g. replica) gets initialized after we indexed the document it gets refreshed internally and we find the doc and its term_vectors, thus the test fails
We currently use the djb2 hash function in order to compute the shard a
document should go to. Unfortunately this hash function is not very
sophisticated and you can sometimes hit adversarial cases, such as numeric ids
on 33 shards.
Murmur3 generates hashes with a better distribution, which should avoid the
adversarial cases.
Here are some examples of how 100000 incremental ids are distributed to shards
using either djb2 or murmur3.
5 shards:
Murmur3: [19933, 19964, 19940, 20030, 20133]
DJB: [20000, 20000, 20000, 20000, 20000]
3 shards:
Murmur3: [33185, 33347, 33468]
DJB: [30100, 30000, 39900]
33 shards:
Murmur3: [2999, 3096, 2930, 2986, 3070, 3093, 3023, 3052, 3112, 2940, 3036, 2985, 3031, 3048, 3127, 2961, 2901, 3105, 3041, 3130, 3013, 3035, 3031, 3019, 3008, 3022, 3111, 3086, 3016, 2996, 3075, 2945, 2977]
DJB: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 900, 900, 900, 900, 1000, 1000, 10000, 10000, 10000, 10000, 9100, 9100, 9100, 9100, 9000, 9000, 0, 0, 0, 0, 0, 0]
Even if djb2 looks ideal in some cases (5 shards), the fact that the
distribution of its hashes has some patterns can raise issues with some shard
counts (eg. 3, or even worse 33).
Some tests have been modified because they relied on implementation details of
the routing hash function.
Close#7954
Adrien Grand
Honza Král
aleph-zero
Michael McCandless
Simon Willnauer
tlrx
Martijn van Groningen
Ryan Ernst
Boaz Leskes
Jun Ohtani
Spencer Alger
javanna
Leonardo Menezes
Christoph Büscher
Clinton Gormley
Tim Schlechter
Lee Hinman
Alexander Reelsen
Andrew Ochsner
David Pilato
Colin Goodheart-Smithe
Alex Ksikes
Martijn Laarman
Sebastian Ziebell
Karel Minarik
nwarz
Chris Mattmann
Veres Lajos