jenkins-bot has submitted this change and it was merged.
Change subject: Rewrite replica handling
......................................................................
Rewrite replica handling
Use auto_expand_replicas. It's way nicer and easier to control.
Change-Id: Ia707b3cabefab5ac5abf002017dc4d4c5f3fd078
---
M CirrusSearch.php
M README
M maintenance/checkIndexes.php
M maintenance/updateOneSearchIndexConfig.php
M maintenance/updateVersionIndex.php
5 files changed, 31 insertions(+), 40 deletions(-)
Approvals:
Manybubbles: Looks good to me, approved
jenkins-bot: Verified
diff --git a/CirrusSearch.php b/CirrusSearch.php
index b1b3364..c9ced2a 100644
--- a/CirrusSearch.php
+++ b/CirrusSearch.php
@@ -45,11 +45,12 @@
// Number of shards for each index
$wgCirrusSearchShardCount = array( 'content' => 4, 'general' => 4 );
-// Number of replicas per shard for each index
-// The default of 0 is fine for single-node setups, but if this is
-// deployed to a multi-node setting you probably at least want these
-// set to 1 for some redundancy, if not 2 for more redundancy.
-$wgCirrusSearchReplicaCount = array( 'content' => 0, 'general' => 0 );
+// Number of replicas Elasticsearch can expand or contract to. This allows for
+// easy development and deployment to a single node (0 replicas) to scale up to
+// higher levels of replication. You if you need more redundancy you could
+// adjust this to '0-10' or '0-all' or even 'false' (string, not boolean) to
+// disable the behavior entirely. The default should be fine for most people.
+$wgCirrusSearchReplicas = '0-2';
// How many seconds must a search of Elasticsearch be before we consider it
// slow? Default value is 10 seconds which should be fine for catching the
rare
@@ -80,8 +81,7 @@
// mapped to specific index suffixes. The keys are the namespace number, and
// the value is a string name of what index suffix to use. Changing this
setting
// requires a full reindex (not in-place) of the wiki. If this setting
contains
-// any values then the index names must also exist in $wgCirrusSearchShardCount
-// and $wgCirrusSearchReplicaCount.
+// any values then the index names must also exist in
$wgCirrusSearchShardCount.
$wgCirrusSearchNamespaceMappings = array();
// Extra indexes (if any) you want to search, and for what namespaces?
diff --git a/README b/README
index af38c63..21d9ef5 100644
--- a/README
+++ b/README
@@ -17,8 +17,6 @@
Configure your search servers in LocalSettings.php if you aren't running
Elasticsearch on localhost:
$wgCirrusSearchServers = array( 'elasticsearch0', 'elasticsearch1',
'elasticsearch2', 'elasticsearch3' );
There are other $wgCirrusSearch variables that you might want to change from
their defaults.
-In production setups, you will very likely want to increase the number of
replicas using
-$wgCirrusSearchReplicaCount.
Now run this script to generate your elasticsearch index:
php
$MW_INSTALL_PATH/extensions/CirrusSearch/maintenance/updateSearchIndexConfig.php
@@ -198,15 +196,6 @@
# Disable ability to shutdown nodes via REST API.
##
action.disable_shutdown: true
-
-
-CirrusSearch
-
-You should change the number of replicas to at least one, two if you want
protection from crashes
-during upgrades. More replicas will help distribute queries across more
nodes. Having more
-replicas than nodes doesn't make sense and Elasticsearch will stay "yellow" if
you do that.
-Anyway, you should add this to LocalSettings.php to change it:
- $wgCirrusSearchReplicaCount = array( 'content' => 2, 'general' => 2 );
Testing
diff --git a/maintenance/checkIndexes.php b/maintenance/checkIndexes.php
index 15739c4..d85df37 100644
--- a/maintenance/checkIndexes.php
+++ b/maintenance/checkIndexes.php
@@ -47,7 +47,7 @@
}
$this->ensureClusterStateFetched();
$this->ensureCirrusInfoFetched();
- $this->checkIndex( 'mw_cirrus_versions', 1, 1 );
+ $this->checkIndex( 'mw_cirrus_versions', 1 );
$aliases = array();
foreach ( $this->clusterState[ 'metadata' ][ 'indices' ] as
$indexName => $data ) {
foreach ( $data[ 'aliases' ] as $alias ) {
@@ -56,7 +56,7 @@
}
foreach ( $this->cirrusInfo as $alias => $data ) {
foreach ( $aliases[ $alias ] as $indexName ) {
- $this->checkIndex( $indexName, $data[
'shard_count'], $data[ 'replica_count' ] );
+ $this->checkIndex( $indexName, $data[
'shard_count'] );
}
}
$indexCount = count( $this->cirrusInfo );
@@ -77,7 +77,7 @@
}
}
- private function checkIndex( $indexName, $expectedShardCount,
$expectedReplicaCount ) {
+ private function checkIndex( $indexName, $expectedShardCount ) {
$this->path = array();
$metdata = $this->getIndexMetadata( $indexName );
$this->in( $indexName );
@@ -92,7 +92,6 @@
$this->check( 'shard count', $expectedShardCount, count(
$routingTable[ 'shards' ] ) );
foreach ( $routingTable[ 'shards' ] as $shardIndex =>
$shardRoutingTable ) {
$this->in( "shard $shardIndex" );
- $this->check( 'replica count', $expectedReplicaCount,
count( $shardRoutingTable ) - 1 );
foreach ( $shardRoutingTable as $replicaIndex =>
$replica ) {
$this->in( "replica $replicaIndex" );
$this->check( 'state', array( 'STARTED',
'RELOCATING' ), $replica[ 'state' ] );
@@ -187,7 +186,6 @@
$data = $r->getData();
$this->cirrusInfo[ $r->getId() ] = array(
'shard_count' => $data[ 'shard_count' ],
- 'replica_count' => $data[
'replica_count' ],
);
}
}
diff --git a/maintenance/updateOneSearchIndexConfig.php
b/maintenance/updateOneSearchIndexConfig.php
index 98d50fc..bf9bd9e 100644
--- a/maintenance/updateOneSearchIndexConfig.php
+++ b/maintenance/updateOneSearchIndexConfig.php
@@ -305,13 +305,13 @@
"validate everything else.", 1 );
}
- $this->output( $this->indent . "\tValidating number of
replicas..." );
- $actualReplicaCount = $settings[ 'number_of_replicas' ];
+ $this->output( $this->indent . "\tValidating replica range..."
);
+ $actualReplicaCount = isset( $settings[ 'auto_expand_replicas'
] ) ? $settings[ 'auto_expand_replicas' ] : 'false';
if ( $actualReplicaCount == $this->getReplicaCount() ) {
$this->output( "ok\n" );
} else {
$this->output( "is $actualReplicaCount but should be "
. $this->getReplicaCount() . '...' );
- $this->getIndex()->getSettings()->setNumberOfReplicas(
$this->getReplicaCount() );
+ $this->getIndex()->getSettings()->set( array(
'auto_expand_replicas' => $this->getReplicaCount() ) );
$this->output( "corrected\n" );
}
}
@@ -550,7 +550,7 @@
}
$this->validateIndexSettings();
$this->output( $this->indent . "\tWaiting for
all shards to start...\n" );
- $expectedActive = $this->getShardCount() * ( 1
+ $this->getReplicaCount() );
+ list( $lower, $upper ) = explode( '-',
$this->getReplicaCount() );
$indexName = $this->getSpecificIndexName();
$path = "_cluster/health/$indexName";
$each = 0;
@@ -566,6 +566,17 @@
$relocating = $health[
'relocating_shards' ];
$initializing = $health[
'initializing_shards' ];
$unassigned = $health[
'unassigned_shards' ];
+ $nodes = $health['number_of_nodes'];
+ if ( $nodes < $lower ) {
+ $this->error( "Require $lower
replicas but only have $nodes nodes. "
+ . "This is almost
always due to misconfiguration, aborting.", 1 );
+ }
+ // If the upper range is all, expect
the upper bound to be the number of nodes
+ if ( $upper === 'all' ) {
+ $upper = $nodes;
+ }
+ $expectedReplicas = min( max( $nodes,
$lower ), $upper );
+ $expectedActive =
$this->getShardCount() * ( 1 + $expectedReplicas );
if ( $each === 0 || $active ===
$expectedActive ) {
$this->output( $this->indent .
"\t\tactive:$active/$expectedActive relocating:$relocating " .
"initializing:$initializing unassigned:$unassigned\n" );
@@ -831,7 +842,7 @@
$this->getIndex()->create( array(
'settings' => array(
'number_of_shards' => $this->getShardCount(),
- 'number_of_replicas' =>
$this->reindexAndRemoveOk ? 0 : $this->getReplicaCount(),
+ 'auto_expand_replicas' =>
$this->getReplicaCount(),
'analysis' =>
$this->analysisConfigBuilder->buildConfig(),
'translog.flush_threshold_ops' => 50000, //
This is supposed to help with bulk index io load.
'index.query.default_field' => 'page.text', //
Since the _all field is disabled, we should query something.
@@ -930,12 +941,8 @@
}
private function getReplicaCount() {
- global $wgCirrusSearchReplicaCount;
- if ( !isset( $wgCirrusSearchReplicaCount[ $this->indexType ] )
) {
- $this->error( 'Could not find a replica count for ' .
$this->indexType . '. Did you add an index to ' .
- '$wgCirrusSearchNamespaceMappings but forget to
add it to $wgCirrusSearchReplicaCount?', 1 );
- }
- return $wgCirrusSearchReplicaCount[ $this->indexType ];
+ global $wgCirrusSearchReplicas;
+ return $wgCirrusSearchReplicas;
}
private function parsePotentialPercent( $str ) {
diff --git a/maintenance/updateVersionIndex.php
b/maintenance/updateVersionIndex.php
index 1857bca..a9b7f32 100644
--- a/maintenance/updateVersionIndex.php
+++ b/maintenance/updateVersionIndex.php
@@ -73,14 +73,13 @@
"{$data['analysis_maj']}.{$data['analysis_min']}\n" .
"{$this->indent} mapping version: " .
"{$data['mapping_maj']}.{$data['mapping_min']}\n" .
- "{$this->indent} shards:
{$data['shard_count']}\n" .
- "{$this->indent} replicas:
{$data['replica_count']}\n"
+ "{$this->indent} shards:
{$data['shard_count']}\n"
);
}
}
private function update( $baseName ) {
- global $wgCirrusSearchShardCount, $wgCirrusSearchReplicaCount;
+ global $wgCirrusSearchShardCount;
$versionType = $this->getType();
$this->output( "{$this->indent}Updating tracking indexes..." );
$docs = array();
@@ -95,7 +94,6 @@
'mapping_maj' => $mMaj,
'mapping_min' => $mMin,
'shard_count' =>
$wgCirrusSearchShardCount[ $type ],
- 'replica_count' =>
$wgCirrusSearchReplicaCount[ $type ],
)
);
}
@@ -108,7 +106,7 @@
if ( !$index->exists() ) {
$this->output( "{$this->indent}Creating tracking
index..." );
$index->create( array( 'number_of_shards' => 1,
- 'number_of_replicas' => 0, ), true );
+ 'auto_expand_replicas' => '0-2', ), true );
$mapping = new \Elastica\Type\Mapping();
$mapping->setType( $index->getType( 'version' ) );
$mapping->setProperties( array(
@@ -117,7 +115,6 @@
'mapping_maj' => array( 'type' => 'long',
'include_in_all' => false ),
'mapping_min' => array( 'type' => 'long',
'include_in_all' => false ),
'shard_count' => array( 'type' => 'long',
'include_in_all' => false ),
- 'replica_count' => array( 'type' => 'long',
'include_in_all' => false ),
) );
$mapping->send();
$this->output( "done\n" );
--
To view, visit https://gerrit.wikimedia.org/r/132463
To unsubscribe, visit https://gerrit.wikimedia.org/r/settings
Gerrit-MessageType: merged
Gerrit-Change-Id: Ia707b3cabefab5ac5abf002017dc4d4c5f3fd078
Gerrit-PatchSet: 5
Gerrit-Project: mediawiki/extensions/CirrusSearch
Gerrit-Branch: master
Gerrit-Owner: Chad <ch...@wikimedia.org>
Gerrit-Reviewer: Chad <ch...@wikimedia.org>
Gerrit-Reviewer: Manybubbles <never...@wikimedia.org>
Gerrit-Reviewer: jenkins-bot <>
_______________________________________________
MediaWiki-commits mailing list
MediaWiki-commits@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/mediawiki-commits
