forward

The out_forward Buffered Output plugin forwards events to other fluentd nodes. This plugin supports load-balancing and automatic fail-over (i.e. active-active backup). For replication, please use out_copy plugin.

The out_forward plugin detects server faults using a φ accrual failure detector algorithm. You can customize the parameters of the algorithm. When a server fault recovers, the plugin makes the server available automatically after a few seconds.

The out_forward plugin supports at-most-once and at-least-once semantics. The default is at-most-once.

It is included in Fluentd's core.

Example Configuration

<match pattern>
  @type forward
  send_timeout 60s
  recover_wait 10s
  hard_timeout 60s

  <server>
    name myserver1
    host 192.168.1.3
    port 24224
    weight 60
  </server>
  <server>
    name myserver2
    host 192.168.1.4
    port 24224
    weight 60
  </server>
  ...

  <secondary>
    @type file
    path /var/log/fluent/forward-failed
  </secondary>
</match>

Please see the Configuration File article for the basic structure and syntax of the configuration file.

Supported Modes

• Synchronous

• Asynchronous

See Output Plugin Overview for more details.

Plugin Helpers

• socket

• server

• timer

• thread

• compat_parameters

Parameters

Common Parameters

@type

The value must be forward.

<server> (at least one is required)

required	multi	version
true	true	0.14.5

The destination servers. Each server has the following parameters:

• host

• name

• port

• shared_key

• username

• password

• standby

• weight

If you want to manage destination servers by flexible approach, use <service_discovery> instead.

host

type	default	version
string	required parameter	0.14.5

The IP address or host name of the server.

name

type	default	version
string	nil	0.14.5

The name of the server. Used for logging and certificate verification in TLS transport (when the host is the address).

port

type	default	version
integer	24224	0.14.5

The port number of the host. Note that both TCP packets (event stream) and UDP packets (heartbeat messages) are sent to this port.

shared_key

type	default	version
string	nil	0.14.5

The shared key per server.

username

type	default	version
string	"" (empty string)	0.14.5

The username for authentication.

password

type	default	version
string	"" (empty string)	0.14.5

The password for authentication.

standby

type	default	version
bool	false	0.14.5

Marks a node as the standby node for an Active-Standby model between Fluentd nodes. When an active node goes down, the standby node is promoted to an active node. The standby node is not used by the out_forward plugin until then.

<match pattern>
  @type forward
  # ...

  <server>
    name myserver1
    host 192.168.1.3
    weight 60
  </server>
  <server>  # forward doesn't use myserver2 until myserver1 goes down
    name myserver2
    host 192.168.1.4
    weight 60
    standby
  </server>
  # ...
</match>

weight

type	default	version
integer	60	0.14.5

The load balancing weight. If the weight of one server is 20 and the weight of the other server is 30, events are sent in a 2:3 ratio. The default weight is 60.

<service_discovery>

Use service discovery plugin instead of fixed <server> list. See also Service Discovery Plugin Overview for more details.

<match pattern>
  @type forward

  <service_discovery>
    @type file
    path /path/to/servers.yaml
  </service_discovery>
</source>

require_ack_response

type	default	version
bool	false	0.14.0

Changes the protocol to at-least-once. The plugin waits for the ack from destination's in_forward plugin.

ack_response_timeout

type	default	version
time	190	0.14.0

This option is used when require_ack_response is true. This default value is based on popular tcp_syn_retries.

If set 0, this plugin does not wait for the ack response.

send_timeout

type	default	version
time	60	0.14.0

The timeout time when sending event logs.

connect_timeout

type	default	version
time	nil(no timeout)	1.6.0

The connection timeout for the socket. When the connection is timed out during the connection establishment, Errno::ETIMEDOUT error is raised.

recover_wait

type	default	version
time	10	0.14.0

The wait time before accepting a server fault recovery.

heartbeat_type

type	default	available	version
enum	transport	transport, tcp, udp, none	0.14.12

Specifies the transport protocol for heartbeats. Set none to disable.

heartbeat_interval

type	default	version
time	1	0.14.0

The interval of the heartbeat packer.

phi_failure_detector

type	default	version
bool	true	0.14.0

Use the "Phi accrual failure detector" to detect server failure.

phi_threshold

type	default	version
integer	16	0.14.0

The threshold parameter used to detect server faults.

phi_threshold is directly related to heartbeat_interval. If you are using longer heartbeat_interval, please use the larger phi_threshold. Otherwise, you will see frequent detachments of destination servers. The default value 16 is tuned for heartbeat_interval 1s.

hard_timeout

type	default	version
time	60	0.14.0

The hard timeout used to detect server failure. The default value is equal to the send_timeout parameter.

expire_dns_cache

type	default	version
time	nil (persistent cache)	0.14.0

Sets TTL to expire DNS cache in seconds. Set 0 not to use DNS Cache.

dns_round_robin

type	default	version
bool	false	0.14.0

Enable client-side DNS round robin. Uniform randomly pick an IP address to send data when a hostname has several IP addresses.

heartbeat_type udp is not available with dns_round_robintrue. Use heartbeat_type tcp or heartbeat_type none.

ignore_network_errors_at_startup

type	default	version
bool	false	0.14.12

Ignores DNS resolution and errors at startup time.

compress

type	default	version
enum	text	0.14.7

Since v0.14.7, Fluentd supports transparent data compression. You can use this feature to reduce the transferred payload size.

Example:

<match debug.**>
  @type forward
  compress gzip
  <server>
    host 192.168.1.2
    port 24224
  </server>
</match>

You do not need any configuration in the receiving server. Data compression is auto-detected and handled transparently by the destination node.

Since v1.19.3, the receiving in_forward limits the decompressed chunk size to 256MB by default. See in_forward if you need to send larger chunks.

Supported values:

• text

• gzip

• zstd (Experimental) (since v1.19.0)

zstd is an experimental feature supported since v1.19.0. Please make sure that the destination server also supports this feature before using it.

• in_forward plugin supports it since v1.19.0.

tls_version

type	default	available	version
enum	TLSv1_2	TLSv1_1, TLSv1_2	0.14.12

The default version of TLS transport.

tls_ciphers

type	default	version
string	ALL:!aNULL:!eNULL:!SSLv2 (OpenSSL > 1.0.0 default)	0.14.12

The cipher configuration of TLS transport.

tls_insecure_mode

type	default	version
bool	false	0.14.12

Skips all verification of certificates or not.

tls_allow_self_signed_cert

type	default	version
bool	false	0.14.12

Allows self-signed certificates or not.

tls_verify_hostname

type	default	version
bool	true	0.14.12

Verifies hostname of servers and certificates or not in TLS transport.

If the following conditions are met, you must set tls_verify_hostname false explicitly to forward events correctly:

• specify host in <server> section with IP address, not hostname

• specify server certificate file with tls_cert_path which contains common name (CN) field with IP address, not hostname

tls_cert_path

type	default	version
array of string	nil	0.14.12

The additional CA certificate path for TLS.

tls_client_cert_path

type	default	version
string	nil	1.3.2

The client certificate path for TLS.

tls_client_private_key_path

type	default	version
string	nil	1.3.2

The client private key path for TLS.

tls_client_private_key_passphrase

type	default	version
string	nil	1.3.2

The TLS private key passphrase for the client.

tls_cert_thumbprint

type	default	version
string	nil	1.7.1

The certificate thumbprint for searching from Windows system certstore. This parameter is for Windows only.

tls_cert_logical_store_name

type	default	version
string	nil	1.7.1

The certificate logical store name on Windows system certstore. This parameter is for Windows only.

tls_cert_use_enterprise_store

type	default	version
string	true	1.7.1

Enables the certificate enterprise store on Windows system certstore. This parameter is for Windows only.

keepalive

type	default	version
bool	false	1.5.0

Enables the keepalive connection.

keepalive_timeout

type	default	version
time	nil	1.5.0

Timeout for keepalive. Default value is nil which means to keep the connection alive as long as possible.

<security> Section

required	multi	version
false	false	0.14.5

This section contains parameters related to authentication:

• self_hostname

• shared_key

self_hostname

type	default	version
string	required parameter	0.14.5

The hostname.

shared_key

type	default	version
string	required parameter	0.14.5

The shared key for authentication. If you want to specify shared_key for specific server, use the <server> section.

<secondary>

required	multi	version
false	false	0.14.0

The backup destination that is used when all servers are unavailable.

For more details, see Secondary Output.

verify_connection_at_startup

type	default	version
bool	false	1.3.1

Verify that a connection can be made with one of out_forward nodes at the time of startup.

<buffer>

See Buffer Section Configurations for more details.

chunk_keys

type	default	version
array	tag	0.14.9

Overwrites the default value in this plugin.

Tips and Tricks

How to connect to a TLS/SSL enabled server?

If you have set up TLS/SSL encryption for the receiving server, you need to tell the output forwarder to use encryption by setting the transport parameter:

<match debug.**>
  @type forward
  transport tls
  <server>
    host example.com
    port 24224
  </server>
</match>

If you are using a self-signed certificate, copy the certificate file to the forwarding server, then add the following settings:

<match debug.**>
  @type forward
  transport tls
  tls_cert_path /path/to/fluentd.crt # Set the path to the certificate file.
  <server>
    # Set the remote server name. This name should match the Common Name
    # field in the certificate.
    host example.com
    port 24224
  </server>
</match>

After updating the settings, please confirm that the forwarded data is being received by the destination node properly.

How to connect to a TLS/SSL enabled server with Windows Certstore Certificate?

If you have set up TLS/SSL encryption in the receiving server, you need to tell the output forwarder to use encryption by setting the transport parameter.

Valid logical store names are:

• MY

• CA

• ROOT

• AUTHROOT

• DISALLOWED

• SPC

• TRUST

• TRUSTEDPEOPLE

• TRUSTEDPUBLISHER

• CLIENTAUTHISSUER

• TRUSTEDDEVICES

• SMARTCARDROOT

• WEBHOSTING

• REMOTE DESKTOP

Logical store name is case-insensitive. Note that these section configurations work only for Windows.

<match debug.**>
  @type forward
  transport tls
  # Set valid logical store name.
  tls_cert_logical_store_name Trust
  tls_verify_hostname true # Set false to ignore the cert hostname.
  <server>
    host 192.168.1.2
    port 24224
  </server>
</match>

If you are using a self-signed certificate, export the certificate file from Windows Certstore and copy to the forwarding server, then add the following settings:

<match debug.**>
  @type forward
  transport tls
  # Set certificate SHA1 hash which is usually called as thumbprint.
  tls_cert_thumbprint <YOUR CERTIFICATE THUMBPRINT>
  tls_cert_logical_store_name Trust
  tls_verify_hostname true # Set false to ignore cert hostname.
  tls_cert_use_enterprise_store true # Set false to use non-enterprise certificate.
  <server>
    host 192.168.1.2
    port 24224
  </server>
</match>

Note that this configuration works for root certificate which is put in Windows Certstore. Currently, the chained certificate is not supported.

Certificate thumbprint is able to obtain with certutil command.

Here is an example which uses a certificate that is generated by fluent-ca-generate command:

PS> certutil -store -enterprise <YOUR CERTIFICATE STORE NAME>
trust "Enterprise Trust"
================ Certificate 0 ================
Serial Number: 01
Issuer: CN=Fluentd Forward CA, L=Mountain View, S=CA, C=US
 NotBefore: 1/1/1970 12:00 AM
 NotAfter: 8/21/2024 7:43 AM
Subject: CN=Fluentd Forward CA, L=Mountain View, S=CA, C=US
Signature matches Public Key
Root Certificate: Subject matches Issuer
Cert Hash(sha1): <YOUR CERTIFICATE THUMBPRINT>
No key provider information
Cannot find the certificate and private key for decryption.
CertUtil: -store command completed successfully.

Cert hash(sha1) is called as thumbprint in this section.

Note that -enterprise flag represents to use enterprise certstore. Please pay attention to using whether enterprise certificates store or not.

How to Enable Password Authentication?

If you want to connect to a server that requires password authentication, you need to set your credentials in the configuration file:

<match debug.**>
  @type forward
  <server>
    host 192.168.1.2
    port 24224
  </server>
  <security>
    self_hostname HOSTNAME
    shared_key secret
  </security>
</match>

Note that, as to the option self_hostname, you need to set the name of the server on which your out_forward instance is running. In the current implementation, it is considered invalid if your in_forward and out_forward share the same hostname.

What is a Phi accrual failure detector?

Fluentd implements an adaptive failure detection mechanism called "Phi accrual failure detector". Here is how it works:

1. Each in_forward node sends heartbeat packets to its out_forward server

   with a regular interval.

2. The out_forward server records the arrival time of heartbeat packets sent

   by each node.

3. If the server does not receive a heartbeat from one of its nodes for "a long

   time", it assumes the node is down.

But how long should the server wait before detaching a node? The phi accrual failure detector answers this question by computing the probability of a node being down based on the assumption that heartbeat intervals follow normal distribution. Internally, it represents the confidence of a node being down by a continuous function φ(t) which grows as the time from the last packet increases.

For example, suppose that the historical average interval is 1 second and the standard deviation is 1, it is unlikely that the node is still being active when its heartbeat has not been received for the last 10 seconds.

For details, please read the original paper: Hayashibara, Naohiro, et al. "The φ accrual failure detector." IEEE, 2004.

Troubleshooting

"no nodes are available"

Please make sure that you can communicate with port 24224 using not only TCP, but also UDP. These commands will be useful for checking the network configuration:

$ telnet host 24224
$ nmap -p 24224 -sU host

Please note that there is one known issue where VMware will occasionally lose small UDP packets used for heartbeat.

If this article is incorrect or outdated, or omits critical information, please let us know. Fluentd is an open-source project under Cloud Native Computing Foundation (CNCF). All components are available under the Apache 2 License.