Goodbye and thank you synced flush! (#50882)

Synced flush was a brilliant idea. It supports instant recoveries with a 
quite small implementation. However, with the presence of sequence
numbers and retention leases, it is no longer needed. This change
removes it from 8.0.

Relates #5077
This commit is contained in:
Nhat Nguyen 2020-01-16 09:43:07 -05:00 committed by GitHub
parent adb56e5842
commit 09b46c8646
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
53 changed files with 290 additions and 3731 deletions

View file

@ -1,62 +0,0 @@
--
:api: flush-synced
:request: SyncedFlushRequest
:response: SyncedFlushResponse
--
[id="{upid}-{api}"]
=== Flush Synced API
[id="{upid}-{api}-request"]
==== Flush Synced Request
A +{request}+ can be applied to one or more indices, or even on `_all` the indices:
["source","java",subs="attributes,callouts,macros"]
--------------------------------------------------
include-tagged::{doc-tests-file}[{api}-request]
--------------------------------------------------
<1> Flush synced one index
<2> Flush synced multiple indices
<3> Flush synced all the indices
==== Optional arguments
["source","java",subs="attributes,callouts,macros"]
--------------------------------------------------
include-tagged::{doc-tests-file}[{api}-request-indicesOptions]
--------------------------------------------------
<1> Setting `IndicesOptions` controls how unavailable indices are resolved and
how wildcard expressions are expanded
include::../execution.asciidoc[]
[id="{upid}-{api}-response"]
==== Flush Synced Response
The returned +{response}+ allows to retrieve information about the
executed operation as follows:
["source","java",subs="attributes,callouts,macros"]
--------------------------------------------------
include-tagged::{doc-tests-file}[{api}-response]
--------------------------------------------------
<1> Total number of shards hit by the flush request
<2> Number of shards where the flush has succeeded
<3> Number of shards where the flush has failed
<4> Name of the index whose results we are about to calculate.
<5> Total number of shards for index mentioned in 4.
<6> Successful shards for index mentioned in 4.
<7> Failed shards for index mentioned in 4.
<8> One of the failed shard ids of the failed index mentioned in 4.
<9> Reason for failure of copies of the shard mentioned in 8.
<10> JSON represented by a Map<String, Object>. Contains shard related information like id, state, version etc.
for the failed shard copies. If the entire shard failed then this returns an empty map.
By default, if the indices were not found, an `ElasticsearchException` will be thrown:
["source","java",subs="attributes,callouts,macros"]
--------------------------------------------------
include-tagged::{doc-tests-file}[{api}-notfound]
--------------------------------------------------
<1> Do something if the indices to be flushed were not found

View file

@ -102,7 +102,6 @@ Index Management::
* <<{upid}-clone-index>>
* <<{upid}-refresh>>
* <<{upid}-flush>>
* <<{upid}-flush-synced>>
* <<{upid}-clear-cache>>
* <<{upid}-force-merge>>
* <<{upid}-rollover-index>>
@ -138,7 +137,6 @@ include::indices/split_index.asciidoc[]
include::indices/clone_index.asciidoc[]
include::indices/refresh.asciidoc[]
include::indices/flush.asciidoc[]
include::indices/flush_synced.asciidoc[]
include::indices/clear_cache.asciidoc[]
include::indices/force_merge.asciidoc[]
include::indices/rollover.asciidoc[]