Introduce CCR getting started guide (#35434)

This commit introduces a basic getting started guide for cross-cluster
replication to the docs.

Co-authored-by: "lcawl <lcawley@elastic.co>"
This commit is contained in:
Jason Tedor 2018-11-13 06:45:00 -05:00 committed by GitHub
parent 76cefb06e3
commit 92fef40dbf
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 324 additions and 2 deletions

View File

@ -1,7 +1,329 @@
[role="xpack"] [role="xpack"]
[testenv="platinum"] [testenv="platinum"]
[[ccr-getting-started]] [[ccr-getting-started]]
== Getting Started == Getting Started with {ccr}
beta[] beta[]
This is the getting started section of the {ccr} docs.
This getting-started guide for {ccr} shows you how to:
* <<ccr-getting-started-remote-cluster,Connect a local cluster to a remote
cluster>>
* <<ccr-getting-started-leader-index,Create a leader index>> in a remote cluster
* <<ccr-getting-started-follower-index,Create a follower index>> that replicates
a leader index
* <<ccr-getting-started-auto-follow,Automatically create follower indices>>
[float]
[[ccr-getting-started-before-you-begin]]
=== Before you begin
. {stack-gs}/get-started-elastic-stack.html#install-elasticsearch[Install {es}]
on your local and remote clusters.
. Obtain a license that includes the {ccr} features. See
https://www.elastic.co/subscriptions[subscriptions] and
<<license-management>>.
. If the Elastic {security-features} are enabled in your local and remote
clusters, you need a user that has appropriate authority to perform the steps
in this tutorial.
+
--
[[ccr-getting-started-security]]
The {ccr} features use cluster privileges and built-in roles to make it easier
to control which users have authority to manage {ccr}.
By default, you can perform all of the steps in this tutorial by
using the built-in `elastic` user. However, a password must be set for this user
before the user can do anything. For information about how to set that password,
see <<security-getting-started>>.
If you are performing these steps in a production environment, take extra care
because the `elastic` user has the `superuser` role and you could inadvertently
make significant changes.
Alternatively, you can assign the appropriate privileges to a user ID of your
choice. On the remote cluster that contains the leader index, a user will need
the `read_ccr` cluster privilege and `monitor` and `read` privileges on the
leader index.
[source,yml]
--------------------------------------------------
ccr_user:
cluster:
- read_ccr
indices:
- names: [ 'leader-index' ]
privileges:
- monitor
- read
--------------------------------------------------
On the local cluster that contains the follower index, the same user will need
the `manage_ccr` cluster privilege and `monitor`, `read`, `write` and
`manage_follow_index` privileges on the follower index.
[source,yml]
--------------------------------------------------
ccr_user:
cluster:
- manage_ccr
indices:
- names: [ 'follower-index' ]
privileges:
- monitor
- read
- write
- manage_follow_index
--------------------------------------------------
If you are managing
<<ccr-getting-started-remote-cluster,connecting to the remote cluster>> via the
cluster update settings API, you will also need a user with the `all` cluster
privilege.
--
[float]
[[ccr-getting-started-remote-cluster]]
=== Connecting to a remote cluster
The {ccr} features require that you
{ref}/modules-remote-clusters.html[connect your local cluster to a remote
cluster]. In this tutorial, we will connect our local cluster to a remote
cluster with the cluster alias `leader`.
[source,js]
--------------------------------------------------
PUT /_cluster/settings
{
"persistent" : {
"cluster" : {
"remote" : {
"leader" : {
"seeds" : [
"127.0.0.1:9300" <1>
]
}
}
}
}
}
--------------------------------------------------
// CONSOLE
// TEST[setup:host]
// TEST[s/127.0.0.1:9300/\${transport_host}/]
<1> Specifies the hostname and transport port of a seed node in the remote
cluster.
You can verify that the local cluster is successfully connected to the remote
cluster.
[source,js]
--------------------------------------------------
GET /_remote/info
--------------------------------------------------
// CONSOLE
// TEST[continued]
The API will respond by showing that the local cluster is connected to the
remote cluster.
[source,js]
--------------------------------------------------
{
"leader" : {
"seeds" : [
"127.0.0.1:9300"
],
"connected" : true, <1>
"num_nodes_connected" : 1, <2>
"max_connections_per_cluster" : 3,
"initial_connect_timeout" : "30s",
"skip_unavailable" : false
}
}
--------------------------------------------------
// TESTRESPONSE
// TEST[s/127.0.0.1:9300/$body.leader.seeds.0/]
// TEST[s/"connected" : true/"connected" : $body.leader.connected/]
// TEST[s/"num_nodes_connected" : 1/"num_nodes_connected" : $body.leader.num_nodes_connected/]
<1> This shows the local cluster is connected to the remote cluster with cluster
alias `leader`
<2> This shows the number of nodes in the remote cluster the local cluster is
connected to.
[float]
[[ccr-getting-started-leader-index]]
=== Creating a leader index
Leader indices require special index settings to ensure that the operations that
need to be replicated are available when the
follower requests them from the leader. These settings are used to enable soft
deletes on the leader index and to control how many soft deletes are retained. A
_soft delete_ occurs whenever a document is deleted or updated. Soft deletes can
be enabled only on new indices created on or after {es} 6.5.0.
In the following example, we will create a leader index in the remote cluster:
[source,js]
--------------------------------------------------
PUT /server-metrics
{
"settings" : {
"index" : {
"number_of_shards" : 1,
"number_of_replicas" : 0,
"soft_deletes" : {
"enabled" : true, <1>
"retention" : {
"operations" : 1024 <2>
}
}
}
},
"mappings" : {
"metric" : {
"properties" : {
"@timestamp" : {
"type" : "date"
},
"accept" : {
"type" : "long"
},
"deny" : {
"type" : "long"
},
"host" : {
"type" : "keyword"
},
"response" : {
"type" : "float"
},
"service" : {
"type" : "keyword"
},
"total" : {
"type" : "long"
}
}
}
}
}
--------------------------------------------------
// CONSOLE
// TEST[continued]
<1> Enables soft deletes on the leader index.
<2> Sets that up to 1024 soft deletes will be retained.
[float]
[[ccr-getting-started-follower-index]]
=== Creating a follower index
Follower indices are created with the {ref}/ccr-put-follow.html[create follower
API]. When you create a follower index, you must reference the
<<ccr-getting-started-remote-cluster,remote cluster>> and the
<<ccr-getting-started-leader-index,leader index>> that you created in the remote
cluster.
[source,js]
--------------------------------------------------
PUT /server-metrics-copy/_ccr/follow
{
"remote_cluster" : "leader",
"leader_index" : "server-metrics"
}
--------------------------------------------------
// CONSOLE
// TEST[continued]
//////////////////////////
[source,js]
--------------------------------------------------
{
"follow_index_created" : true,
"follow_index_shards_acked" : true,
"index_following_started" : true
}
--------------------------------------------------
// TESTRESPONSE
//////////////////////////
Now when you index documents into your leader index, you will see these
documents replicated in the follower index. You can
inspect the status of replication using the
{ref}/ccr-get-follow-stats[get follower stats API].
//////////////////////////
[source,js]
--------------------------------------------------
POST /server-metrics-copy/_ccr/pause_follow
POST /server-metrics-copy/_close
POST /server-metrics-copy/_ccr/unfollow
--------------------------------------------------
// CONSOLE
// TEST[continued]
//////////////////////////
[float]
[[ccr-getting-started-auto-follow]]
=== Automatically create follower indices
The auto-follow feature in {ccr} helps for time series use cases where you want
to follow new indices that are periodically created in the remote cluster
(such as daily Beats indices). Auto-following is configured using the
{ref}/ccr-put-auto-follow-pattern.html[create auto-follow pattern API]. With an
auto-follow pattern, you reference the
<<ccr-getting-started-remote-cluster,remote cluster>> that you connected your
local cluster to. You must also specify a collection of patterns that match the
indices you want to automatically follow.
For example:
[source,js]
--------------------------------------------------
PUT /_ccr/auto_follow/beats
{
"remote_cluster" : "leader",
"leader_index_patterns" :
[
"metricbeat-*", <1>
"packetbeat-*" <2>
],
"follow_index_pattern" : "{{leader_index}}-copy" <3>
}
--------------------------------------------------
// CONSOLE
// TEST[continued]
<1> Automatically follow new {metricbeat} indices.
<2> Automatically follow new {packetbeat} indices.
<3> The name of the follower index is derived from the name of the leader index
by adding the suffix `-copy` to the name of the leader index.
//////////////////////////
[source,js]
--------------------------------------------------
{
"acknowledged" : true
}
--------------------------------------------------
// TESTRESPONSE
//////////////////////////
//////////////////////////
[source,js]
--------------------------------------------------
DELETE /_ccr/auto_follow/beats
--------------------------------------------------
// CONSOLE
// TEST[continued]
//////////////////////////