Commit Graph

22338 Commits

Author SHA1 Message Date
javanna 1d06916b07 adapt params argument, they can only be Strings in performRequest method 2016-06-03 16:01:07 +02:00
javanna 3890842fd4 add project substitution for org.elasticsearch:client so that we can add it as a dep in other projects 2016-06-03 16:01:07 +02:00
javanna 2735897b36 use versions from versions.properties 2016-06-03 16:01:07 +02:00
javanna 2d7a781195 fix usage of deprecated apis 2016-06-03 16:01:07 +02:00
javanna c0a72c1686 add support for headers: default ones and per request 2016-06-03 16:01:07 +02:00
javanna 85a7721185 simplify Connection class
The connection class can be greatly simplified now that we don't ping anymore. Pings required a special initial state (UNKNOWN) for connections, to indicate that they require pinging although they are not dead. At this point we don't need the State enum anymore, as connections can only be dead or alive based on the number of failed attempts. markResurrected is also not needed, as it was again a way to make pings required. RestClient can simply pick a dead connection now and use it, no need to change its state when picking the connection.
2016-06-03 16:01:07 +02:00
javanna cdffc3d15b remove notion of connection pool, turn around dependency between RestClient and Sniffer
RestClient exposes a setNodes method, which Sniffer can call to update its nodes.
2016-06-03 16:01:07 +02:00
javanna b38ef345e2 remove streams leftover
Given that we don't use streams anymore, we can check straightaway if the connection iterator is empty before returning it and resurrect a connection when needed directly in the connection pool, no lastResortConnection method required.
2016-06-03 16:01:07 +02:00
javanna 599dad560c remove streams and java 8 only api, build with source and target 1.7 2016-06-03 16:01:07 +02:00
javanna 9569ebc262 add builders for simple creation of RestClient and SniffingConnectionPool instances
We have quite some constructor parameters and some defaults should be applied, builders help simplifying creation of objects for users.
2016-06-03 16:01:07 +02:00
javanna e040d2fc77 remove ConnectionPool interface 2016-06-03 16:01:07 +02:00
javanna 17a21f0272 add curl format trace logging for requests and responses 2016-06-03 16:01:07 +02:00
javanna e7fe397c39 add missing parentheses 2016-06-03 16:01:07 +02:00
javanna 530ad227a2 prevent unclosed response entities in ElasticsearchResponseException, eagerly read response string in case of error status code 2016-06-03 16:01:07 +02:00
javanna e77ab87926 return the response as part of ElasticsearchResponseException 2016-06-03 16:01:07 +02:00
javanna e85ed3eb52 remove pinging from static connection pool, can be replaced by a low connect timeout on each request 2016-06-03 16:01:07 +02:00
javanna d7c41764f2 add some javadocs to connection pool classes 2016-06-03 16:01:07 +02:00
javanna 9ffdea9515 remove Scheme enum 2016-06-03 16:01:07 +02:00
javanna bd29dc1572 Remove Transport class, move all of it within RestClient class 2016-06-03 16:01:07 +02:00
javanna 062a21678c merge Connection and StatefulConnection into one class, remove generic type from Transport 2016-06-03 16:01:07 +02:00
javanna a472544ab4 move sniff related stuff to sniff package 2016-06-03 16:01:07 +02:00
javanna ce663e9703 get rid of connection selector predicate 2016-06-03 16:01:07 +02:00
javanna f6a5a0a4ad get rid of Node abstraction 2016-06-03 16:01:07 +02:00
javanna 94cf8437d0 get rid of retry timeout exception 2016-06-03 16:01:07 +02:00
javanna 25892351e7 remove Verb enum 2016-06-03 16:01:07 +02:00
javanna 6ae8be55bd Introduce notion of Transport, Connection and ConnectionPool
There are two implementations of connection pool, a static one that allows to enable/disable pings, and a sniffing one that sniffs nodes from the nodes info api.

Transport retrieves a stream of connections from the connection for each request and calls onSuccess or onFailure depending on the result of the request.

Transport also supports a max retry timeout to control the timeout for the request retries overall.
2016-06-03 16:01:07 +02:00
javanna 5970723945 clarify why jarhell and dependency license check are disabled, add okhttp mockwebserver dependency
Also enable forbiddenApisTest (will fail till we get rid of the sun http web server in RestClientTests that will be replaced with mockwebserver)
2016-06-03 16:01:07 +02:00
javanna c9c33a7719 Fix checkstyle issues, setup thirdPartyAudit checks and enable forbiddenApisMain 2016-06-03 16:01:07 +02:00
Simon Willnauer 3efbe95ca4 RestClient prototype 2016-06-03 16:01:07 +02:00
Yannick Welsch 24a7b7224b Fix recovery throttling to properly handle relocating non-primary shards (#18701)
Relocation of non-primary shards is realized by recovering from the primary shard. Recovery throttling wrongly equates non-primary relocation as recovering a shard from the non-primary relocation source, however.

Closes #18640
2016-06-03 14:11:34 +02:00
Simon Willnauer 6c28235b03 Fix merge stats rendering in RestIndicesAction (#18720)
give the table description:
```
        table.addCell("merges.total", "sibling:pri;alias:mt,mergesTotal;default:false;text-align:right;desc:number of completed merge ops");
        table.addCell("pri.merges.total", "default:false;text-align:right;desc:number of completed merge ops");

        table.addCell("merges.total_docs", "sibling:pri;alias:mtd,mergesTotalDocs;default:false;text-align:right;desc:docs merged");
        table.addCell("pri.merges.total_docs", "default:false;text-align:right;desc:docs merged");

        table.addCell("merges.total_size", "sibling:pri;alias:mts,mergesTotalSize;default:false;text-align:right;desc:size merged");
        table.addCell("pri.merges.total_size", "default:false;text-align:right;desc:size merged");
```
this is how it should be.
2016-06-03 13:42:09 +02:00
Britta Weber 78574d248c [TEST] mute RandomAllocationDeciderTests.testRandomDecisions
we have a pr already: https://github.com/elastic/elasticsearch/pull/18701
2016-06-03 11:37:22 +02:00
Clinton Gormley e6aaaf11ed Reworked docs for index-shrink API (#18705) 2016-06-03 09:50:51 +02:00
Robert Muir ad118eb1e5 Merge pull request #18711 from rmuir/painless_compile_exceptions
Improve painless compile-time exceptions
2016-06-02 20:43:01 -04:00
Robert Muir b5393ce55e Improve painless compile-time exceptions 2016-06-02 18:23:47 -04:00
Ali Beyad b720216395 Adds UUIDs to snapshots
This commit adds a UUID for each snapshot, in addition to the already
existing repository and snapshot name. The addition of UUIDs will enable
more robust handling of the deletion of previous snapshots and lingering
files from partially failed delete operations, on top of being able to
uniquely track each snapshot.

Closes #18228
Relates #18156
2016-06-02 17:01:48 -04:00
Nik Everett f82ab787a5 Add test rethrottle test case for delete-by-query
and remove some type parameters that we don't need that were getting
in the way.
2016-06-02 15:04:18 -04:00
Jason Tedor f67e5807ca Do not start scheduled pings until transport start
Today, scheduled pings in NettyTransport can start before the transport
is started. Instead, these pings should not be scheduled until after the
transport is started. This commit modifies NettyTransport so that this
is the case.

Relates #18702
2016-06-02 13:12:21 -04:00
Christoph Büscher b6f26c0c37 Merge pull request #17624
Handle empty query bodies at parse time and remove EmptyQueryBuilder.
2016-06-02 18:40:36 +02:00
Christoph Büscher 9067407cdd Adressing review comments 2016-06-02 16:19:23 +02:00
Simon Willnauer 22dfc41521 Only filter intial recovery (post API) when shrinking an index (#18661)
Today we use `index.routing.allocation.include._id` to filter the allocation
for the shrink target index. That has the sideeffect that the user has to
delete that setting / change it once the primary has been recovered (shrink is done)
This PR adds a dedicated filter that can only be set internally that only filters
allocation for unassigned shards.
2016-06-02 15:38:51 +02:00
Christoph Büscher e2b6dbc020 Add tests to check that toQuery() doesn't return null 2016-06-02 11:25:56 +02:00
Christoph Büscher cb145aec68 Removing handling of null lucene query where we catch this at parse time 2016-06-02 11:25:56 +02:00
Christoph Büscher 359f45988f Handle empty query bodies at parse time and remove EmptyQueryBuilder
Currently we support empty query clauses like the filter in

"constant_score" : {  "filter" : { } }

How these clauses are handled depends on the surrounding query.
They later are either ignored, converted to match all or no documents or
passed up further in the query hierarchy. During parsing these claues are
currently represented as EmptyQueryBuilders. When not handled anywhere else,
these special cases need to be checked for on the shard when building the
lucene query.

This is trappy, so this PR changes the parsing of compound queries. Instead
of returning QueryBuilder, the core query parsing method
QueryShardContext#parseInnerQueryBuilder() now return an Optional which can
be empty in the case of empty query clauses. This has the advantage of forcing
callers to deal with this sooner or later. When encountering empty Optionals,
compound query builders now have the choice to ignore them, pass them on or
rewrite to a different query, depending on context.
2016-06-02 11:25:56 +02:00
Yannick Welsch b2724c0d08 Mute failing assertions in IndexWithShadowReplicasIT until fix 2016-06-02 11:23:44 +02:00
Jason Tedor 3ccd59592a Remove allow running as root
This commit removes the escape hatch for running Elasticsearch as root.

Relates #18694
2016-06-02 05:03:28 -04:00
Clinton Gormley 56d86ef875 Add upgrade-not-supported warning to alpha release notes 2016-06-02 10:18:16 +02:00
Nicholas Knize 2d57bbdacd remove unrecognized javadoc tag from matrix aggregation module 2016-06-01 17:30:20 -05:00
Nicholas Knize 54575e55ca set ValuesSourceConfig fields as private 2016-06-01 16:39:42 -05:00
Nicholas Knize 90b8f5d0d8 Adding MultiValuesSource support classes and documentation to matrix stats agg module 2016-06-01 16:39:42 -05:00