2019-03-13 15:12:52 +00:00
---
2020-04-07 18:55:19 +00:00
layout: docs
page_title: Bootstrapping a Datacenter
description: >-
An agent can run in both client and server mode. Server nodes are responsible
for running the consensus protocol and storing the cluster state. Before a
Consul cluster can begin to service requests, a server node must be elected
leader. Thus, the first nodes that are started are generally the server nodes.
Bootstrapping is the process of joining these server nodes into a cluster.
2019-03-13 15:12:52 +00:00
---
# Bootstrapping a Datacenter
An agent can run in either client or server mode. Server nodes are responsible for running the
2020-04-09 23:46:54 +00:00
[consensus protocol](/docs/internals/consensus) and storing the cluster state.
2019-03-13 15:12:52 +00:00
The client nodes are mostly stateless and rely heavily on the server nodes.
2020-04-06 20:27:35 +00:00
Before a Consul cluster can begin to service requests,
2019-03-13 15:12:52 +00:00
a server node must be elected leader. Bootstrapping is the process
2020-04-06 20:27:35 +00:00
of joining these initial server nodes into a cluster. Read the
2020-04-09 23:46:54 +00:00
[architecture documentation](/docs/internals/architecture) to learn more about
2019-03-13 15:12:52 +00:00
the internals of Consul.
It is recommended to have three or five total servers per datacenter. A single server deployment is _highly_ discouraged
as data loss is inevitable in a failure scenario. Please refer to the
2020-04-09 23:46:54 +00:00
[deployment table](/docs/internals/consensus#deployment-table) for more detail.
2019-03-13 15:12:52 +00:00
~> **Note**: In versions of Consul prior to 0.4, bootstrapping was a manual process. For details on using the `-bootstrap` flag directly, see the
2020-04-09 23:46:54 +00:00
[manual bootstrapping documentation](/docs/install/bootstrapping#manually-join-the-servers).
2020-04-06 20:27:35 +00:00
Manual bootstrapping with `-bootstrap` is not recommended in
newer versions of Consul (0.5 and newer) as it is more error-prone.
2019-03-13 15:12:52 +00:00
Instead you should use automatic bootstrapping
2020-04-09 23:46:54 +00:00
with [`-bootstrap-expect`](/docs/agent/options#_bootstrap_expect).
2019-03-13 15:12:52 +00:00
## Bootstrapping the Servers
2020-04-09 23:46:54 +00:00
The recommended way to bootstrap the servers is to use the [`-bootstrap-expect`](/docs/agent/options#_bootstrap_expect)
2019-03-13 15:12:52 +00:00
configuration option. This option informs Consul of the expected number of
server nodes and automatically bootstraps when that many servers are available. To prevent
inconsistencies and split-brain (clusters where multiple servers consider
themselves leader) situations, you should either specify the same value for
2020-04-09 23:46:54 +00:00
[`-bootstrap-expect`](/docs/agent/options#_bootstrap_expect)
2019-03-13 15:12:52 +00:00
or specify no value at all on all the servers. Only servers that specify a value will attempt to bootstrap the cluster.
Suppose we are starting a three server cluster. We can start `Node A`, `Node B`, and `Node C` with each
2020-04-06 20:27:35 +00:00
providing the `-bootstrap-expect 3` flag. Once the nodes are started, you should see a warning message in the service output.
2019-03-13 15:12:52 +00:00
```text
[WARN] raft: EnableSingleNode disabled, and no known peers. Aborting election.
```
2020-04-06 20:27:35 +00:00
The warning indicates that the nodes are expecting 2 peers but none are known yet. Below you will learn how to connect the servers so that one can be
2019-03-13 15:12:52 +00:00
elected leader.
## Creating the Cluster
You can trigger leader election by joining the servers together, to create a cluster. You can either configure the nodes to join automatically or manually.
### Automatically Join the Servers
There are multiple options for joining the servers. Choose the method which best suits your environment and specific use case.
- Specify a list of servers with
2020-04-09 23:46:54 +00:00
[-join](/docs/agent/options#_join) and
[start_join](/docs/agent/options#start_join)
2019-03-13 15:12:52 +00:00
options.
2020-04-09 23:46:54 +00:00
- Specify a list of servers with [-retry-join](/docs/agent/options#_retry_join) option.
- Use automatic joining by tag for supported cloud environments with the [-retry-join](/docs/agent/options#_retry_join) option.
2019-03-13 15:12:52 +00:00
2020-04-06 20:27:35 +00:00
All three methods can be set in the agent configuration file or
the command line flag.
2019-03-13 15:12:52 +00:00
### Manually Join the Servers
To manually create a cluster, you should connect to one of the servers
and run the `consul join` command.
2020-05-19 18:32:38 +00:00
```shell-session
2019-03-13 15:12:52 +00:00
$ consul join <Node A Address> <Node B Address> <Node C Address>
Successfully joined cluster by contacting 3 nodes.
```
Since a join operation is symmetric, it does not matter which node initiates it. Once the join is successful, one of the nodes will output something like:
2020-04-07 23:56:08 +00:00
```text
2019-03-13 15:12:52 +00:00
[INFO] consul: adding server foo (Addr: 127.0.0.2:8300) (DC: dc1)
[INFO] consul: adding server bar (Addr: 127.0.0.1:8300) (DC: dc1)
[INFO] consul: Attempting bootstrap with nodes: [127.0.0.3:8300 127.0.0.2:8300 127.0.0.1:8300]
...
[INFO] consul: cluster leadership acquired
```
### Verifying the Cluster and Connect the Clients
2020-10-14 15:23:05 +00:00
As a sanity check, the [`consul info`](/commands/info) command
2020-04-06 20:27:35 +00:00
is a useful tool. It can be used to verify the `raft.num_peers`
2019-03-13 15:12:52 +00:00
and to view the latest log index under `raft.last_log_index`. When
2020-10-14 15:23:05 +00:00
running [`consul info`](/commands/info) on the followers, you
2019-03-13 15:12:52 +00:00
should see `raft.last_log_index` converge to the same value once the
leader begins replication. That value represents the last log entry that
has been stored on disk.
2020-04-06 20:27:35 +00:00
Now that the servers are all started and replicating to each other, you can
join the clients with the same join method you used for the servers.
2019-03-13 15:12:52 +00:00
Clients are much easier as they can join against any existing node. All nodes participate in a gossip
protocol to perform basic discovery, so once joined to any member of the
cluster, new clients will automatically find the servers and register
themselves.
-> **Note:** It is not strictly necessary to start the server nodes before the clients; however, most operations will fail until the servers are available.
