MigratoryData Server
Configuration Guide
Version 5.0
March 31, 2023
Copyright © 2007-2021 Migratory Data Systems. ALL RIGHTS RESERVED.
THIS DOCUMENT IS PROVIDED “AS IS” WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, OR NON-INFRINGEMENT.
THIS DOCUMENT COULD INCLUDE TECHNICAL INACCURACIES OR TYPOGRAPHICAL ERRORS. CHANGES ARE PERIODICALLY ADDED TO THE INFORMATION HEREIN; THESE CHANGES WILL BE INCORPORATED IN NEW EDITIONS OF THE DOCUMENT. MIGRATORY DATA SYSTEMS MAY MAKE IMPROVEMENTS AND/OR CHANGES IN THE PRODUCT DESCRIBED IN THIS DOCUMENT AT ANY TIME.
This guide describes the configuration of MigratoryData Server. It is recommended to read MigratoryData Architecture Guide before reading this document for a better understanding of the concepts and to have a more comprehensive background.
This guide is part of the documentation set for MigratoryData Server version 5.0.
The configuration file of the MigratoryData server has comments and optional parameters besides required parameters. The optional parameters have default values. An optional parameter that is not present in the configuration file will be used with its default value.
The configuration file supports comments. A line that starts with a # character is interpreted as a comment and is ignored. A parameter can be configured with the following syntax:
parameter = value
The value of a parameter can be defined on multiple lines by postfixing each line with \. For example
LicenseKey = aaaabbbbcccc
can be written as follows:
LicenseKey = aaaa\
bbbb\
cccc
The configurable parameters are described in Chapter 4 and Chapter 5.
You might use one or more of the environment variables defined below to customize various aspects of your MigratoryData deployment. These environment variables should be defined in one of the following files, which is part of the rpm/deb distribution of the MigratoryData server:
Configuration file for extra options | Platform |
/etc/default/migratorydata | For deb-based Linux |
/etc/sysconfig/migratorydata | For rpm-based Linux |
Every configurable parameter of the MigratoryData server (see Chapter 4 and Chapter 5) can be either configured in the configuration file as described in Chapter 2, or specified as an extra option using the MIGRATORYDATA_EXTRA_OPTS environment variable using the following syntax:
-Dparameter=value
The value of a parameter should be defined without spaces and quotes. For example, use -DMemory=512MB, rather than -DMemory='512 MB', or use -DLogFolder=/tmp/somefile, rather than -DLogFolder="/tmp/some file".
Also, the value cannot be defined on multiple lines. For example, if the LicenseKey parameter is configured in your configuration file on multiple lines as follows:
LicenseKey = aaaa\
bbbb\
cccc
then, you could redefine it as an extra option as follows:
-DLicenseKey=aaaabbbbcccc
A parameter defined using an extra option will override the value of that parameter defined in the configuration file. This is typically the goal of the extra options. For example, an extra option could be used to temporarily test a feature of your deployment without modifying the configuration file. Also, extra options are useful when running the MigratoryData server into a container. In this case, the default configuration file cannot be typically edited inside the container, so customization is done using extra options.
MIGRATORYDATA_EXTRA_OPTS | |
Description | Specifies various extra options for the MigratoryData server which can be used to define or override parameters of the configuration file |
Default value | "" |
Required parameter | Optional |
For example, to define JMX monitoring on the port 3000 and without using authentication, or to alter the JMX monitoring configured using the configuration file, you could configure the MIGRATORYDATA_EXTRA_OPTS environment variable as follows:
MIGRATORYDATA_EXTRA_OPTS='-DMonitor=JMX -DMonitorJMX.Listen=*:3000 -DMonitorJMX.Authentication=false'
Use this environment variable to define various Java options.
MIGRATORYDATA_JAVA_EXTRA_OPTS | |
Description | Specifies various extra options for Java |
Default value | "" |
Required parameter | Optional |
For most cases, you don't need to configure this environment variable. However, there are situations when adding an extra option for Java could be necessary. For example, if you run the MigratoryData server into a docker container or AWS EC2 machine, then, in order to be able to access the JMX monitoring using the jconsole tool, you will need to provide extra options for Java as follows:
MIGRATORYDATA_JAVA_EXTRA_OPTS='-Djava.net.preferIPv4Stack=true -Djava.rmi.server.hostname=yourhostname'
where yourhostname is the DNS name of the machine hosting the docker container, or the public DNS name of the AWS EC2 machine.
Use this environment variable to define the Java options related to Garbage Collection. These options should not include the Java options for Garbage Collection logging which should be provided using the environment variable MIGRATORYDATA_JAVA_GC_LOG_OPTS.
MIGRATORYDATA_JAVA_GC_OPTS | |
Description | Specifies various options for Java Garbage Collection. |
Default value | "-verbosegc -XX:+UseParallelOldGC" |
Required parameter | Optional |
For most cases, it is sufficient to use the default value of this environment variable.
Use this environment variable to define the Java options related to Garbage Collection logging.
MIGRATORYDATA_JAVA_GC_LOG_OPTS | |
Description | Specifies options for Java Garbage Collection logs |
Default value | "-XX:+PrintGCDetails -XX:+PrintGCTimeStamps -XX:+PrintGCDateStamps -XX:+UseGCLogFileRotation -XX:GCLogFileSize=10M -XX:NumberOfGCLogFiles=5 -Xloggc:YourLogFolder/all/gc-%t.log" for Java 8 |
"-Xlog:gc*:file=YourLogFolder/all/gc-%t.log:time, uptimemillis:filecount=5,filesize=10000" for Java 9+ | |
Required parameter | Optional |
For most cases, it is sufficient to use the default value of this environment variable.
Use this environment variable to define the maximum number of socket descriptors to be used by the maximum number of concurrent users connecting to the MigratoryData server.
MIGRATORYDATA_MAX_DESCRIPTORS | |
Description | Specifies the maximum number of socket descriptors. |
Default value | 1000000 |
Required parameter | Optional |
For most cases, it is sufficient to use the default value of this environment variable.
The basic parameters of MigratoryData Server are described below.
LicenseKey | |
Description | A string representing the license key |
Default value | No default value |
Required parameter | Required |
The license key consists of a sequence of numbers and letters. License keys are provided by MigratoryData to customers either for evaluation, development, or production usage of MigratoryData Server. To obtain a valid license key, contact MigratoryData at support@migratorydata.com
Memory | |
Description | The maximum memory for the Java Virtual Machine (JVM) process running MigratoryData Server. The memory can be also expressed in gigabytes, megabytes, and kilobytes by using the following postfixes: GB for gigabytes, MB for megabytes, or KB for kilobytes. For example to allocate 512 megabytes for MigratoryData Server, the parameter Memory should be configured as follows:
Memory = 512 MB |
Default value | No default value |
Required parameter | Required |
In a production environment is is recommended to use at least 8192 MB memory space or more depending on the load of data on the server and how much simultaneous clients will be connected. Use MigratoryData Benchmark Kit to estimate the memory required for your use case.
Listen | |
Description | A comma separated list of addresses (IP address or DNS domain name and its corresponding port) to listen for client connections |
Default value | No default value |
Required parameter | Optional |
The format of the addresses is "IP:Port" (e.g. 192.168.1.1:80) or "Name:Port" (e.g. push.example.com:80). IPv6 addresses must be enclosed in square brackets
(e.g. [2001:db8::a00:20ff:fea7:ccea]:80).
If you specify an address without a port, the default port 80 will be used.
By specifying an IP address, MigratoryData Server will bind only to that IP address. The wildcard address (*) will enable MigratoryData Server to bind to all available IP addresses of the machine.
If MigratoryData Server is installed on a machine that has a firewall enabled, then you will need to allow in firewall the access to the addresses and ports configured by this parameter.
Note -- For web applications, in a production environment, MigratoryData Server requires a standard web server - such as Apache - to serve the static pages, the images, and the server scripting of the web application. The DNS name of the address provided by this Listen parameter must share the same sub-domain with the DNS name used to access the standard web server.
For example, if your Apache web server is accessible at www.example.com, then an example of valid configuration for MigratoryData Server will be:
Listen = push.example.com:80.
but the following configuration will not be valid:
Listen = push.example.org:80.
because the sub-domain example.com of Apache is different from example.org, which is the sub-domain configured for MigratoryData Server in this example.
ListenEncrypted | |
Description | A comma separated list of addresses (IP address or DNS domain name and its corresponding port) to listen for encrypted client connections |
Default value | No default value |
Required parameter | Optional |
The same conventions applies as for the Listen parameter described in Section 4.3, except the fact that if you specify an address without a port, the default port 443 will be used.
Note -- In a production deployment, it is recommend to use HTTPS encrypted client connections. In addition to the fact that data will be securely delivered to clients, there is another important advantage. Some security solutions consider HTTP streaming as a non-standard HTTP access and may block streaming to the client having installed such a security solution. By deploying MigratoryData Server with HTTPS encrypted client connections, the security solutions cannot inspect the content of the HTTP messages so that streaming cannot be blocked.
KeyStore | |
Description | The file containing the entries with SSL certificates for the encrypted connections |
Default value | No default value |
Required parameter | Required if at least one of the following parameters are configured:
|
The keystore file must be configured using absolute paths. A value example for this parameter is as follows:
Example of KeyStore | Platform |
KeyStore = /some/path/mykeystore.jks | For Linux/Unix |
KeyStore = C:/some/path/mykeystore.jks | For Windows |
The keystore must contain a SSL certificate for each address used in the configuration of the following parameters:
Suppose the following DNS address push.example.com resolves to the IP address 192.168.1.1, and vice-versa, the IP address 192.168.1.1 resolves to the DNS address push.example.com.
If an address appears in the configuration of any of the parameters used to define encrypted connections (see Table 4.1), then the keystore file must contain an SSL certificate for that address. If the address is specified by its DNS name, say push.example.com, then its certificate entry in the keystore must have as alias the string push.example.com. If the address is specified by its IP address, say 192.168.1.1, then its certificate entry in the keystore must have as alias the string 192.168.1.1.
To create a certificate for an address in the keystore, there are two possibilities:
Tip -- Self-signed certificates can be used in development. In a production environment always use certificates signed by a Certificate Authority.
In the next two subsections we suppose the keystore file is named mykeystore.jks and the address for which a certificate should be added in the keystore is specified either by DNS name as push.example.com or by IP address as 192.168.1.1.
Enter one of following commands on a single line without line breaks depending on how was specified the address in the configuration file:
Address is used in configuration by DNS name as push.example.com |
keytool -genkeypair -dname "CN=push.example.com" -alias push.example.com -keyalg RSA -validity 3600 -keystore mykeystore.jks |
Address is used in configuration by IP address as 192.168.1.1 |
keytool -genkeypair -dname "CN=192.168.1.1" -alias 192.168.1.1 -keyalg RSA -validity 3600 -keystore mykeystore.jks |
Address is used in configuration for encrypted JMX monitoring |
keytool -genkeypair -dname "CN=jmx" -alias jmx -keyalg RSA -validity 3600 -keystore mykeystore.jks |
You will be asked to set a password for the keystore if the file mykeystore.jks does not exist, or to enter the keystore password if the keystore file already exists and contains other certificate entries. This password must be used to configure the parameter KeyStorePassword (see Section 4.6).
Suppose you obtained the following signed certificate file push.example.com.crt from a Certificate Authority for the domain push.example.com and suppose push.example.com.key is the file containing its corresponding private key.
In order to add this CA-signed certificate to the keystore, follow these steps:
To chain your signed certificate push.example.com.crt with the intermediary certificate, you should first append to your signed certificate the intermediary certificate (say intermediary.crt), and then append the certificate signing request sent by you to the CA authority (say push.example.com.csr). On a Unix-like operating system, use the following command:
cat intermediary.crt push.example.com.csr » push.example.com.crt
On Windows, use the command:
type intermediary.crt push.example.com.csr » push.example.com.crt
openssl
pkcs12
-export
-in push.example.com.crt
-inkey push.example.com.key
-out push.example.com.pkcs12
This will produce the file push.example.com.pkcs12; the password for the this new file must be the same like that used for the keystore.
keytool
-importkeystore
-srckeystore push.example.com.pkcs12
-srcstoretype PKCS12
-deststoretype JKS
-destkeystore mykeystore.jks
This will insert a new entry in the keystore file mykeystore.jks with the default alias 1.
Address is used in configuration by DNS name as push.example.com |
keytool -v -keystore mykeystore.jks -changealias -alias 1 -destalias push.example.com |
Address is used in configuration by IP address as 192.168.1.1 |
keytool -v -keystore mykeystore.jks -changealias -alias 1 -destalias 192.168.1.1 |
Address is used in configuration for encrypted JMX monitoring |
keytool -v -keystore mykeystore.jks -changealias -alias 1 -destalias jmx |
KeyStorePassword | |
Description | The password to access the keystore |
Default value | No default value |
Required parameter | Required if the parameter KeyStore is configured |
Configure this parameter with the password used when you created the keystore file. The keystore file is created when adding the first certificate entry to it as explained in Section 4.5.1.
Monitor | |
Description | Monitoring type (possible values: JMX, HTTP, PUSH) |
Default value | No default value |
Required parameter | Optional |
You can configure JMX monitoring, HTTP monitoring, PUSH monitoring or any combination of them. Use comma separated values to configure all monitoring types as follows:
Monitor = JMX, HTTP, PUSH
PUSH monitoring consists in obtaining monitoring information in real-time using any MigratoryData Client API by subscribing to certain special subjects named meta-subjects. Access to meta-subjects is subject to accessibility, entitlement, and encryption like for any other subjects. Therefore, unlike for JMX and HTTP monitoring, there are no specific parameters for PUSH monitoring concerning accessibility, entitlement, and encryption.
The following meta-subjects are available for PUSH monitoring provided that the parameter MonitorPUSH.Stats is set on true:
Also, the following optional meta-subjects exist:
MonitorUsername | |
Description | User name for monitoring access |
Default value | admin |
Required parameter | Required if at least one of the following parameters is configured:
|
MonitorPassword | |
Description | Password for monitoring access |
Default value | pass |
Required parameter | Required if at least one of the following parameters is configured:
|
MonitorJMX.Listen | |
Description | Address used to listen for JMX compatible clients |
Default value | No default value |
Required parameter | Optional |
The format of this addresse is "IP:Port" (e.g. 192.168.1.1:3000) or "Name:Port" (e.g. push.example.com:3000). IPv6 addresses must be enclosed in square brackets
(e.g. [2001:db8::a00:20ff:fea7:ccea]:3000).
Note -- In order to access the JMX monitoring in MigratoryData Server from a remote machine, please check your firewall settings so that the network traffic from the remote machine is allowed to the address configured by this parameter MonitorJMX.Listen.
The jconsole utility that is freely available with OpenJDK can be used to connect to the JMX monitoring service of MigratoryData Server. Also there are many JMX commercial tools that provide enhanced functionality like dashboards and database persistence that can be used to connect to the JMX monitoring service of MigratoryData Server.
Caution -- The jconsole utility has a known issue when connecting remotely to a JMX service running on Linux. To avoid this issue, execute the following command on the Linux machine where MigratoryData Server runs:
hostname -i
The command above should return the address used in the configuration of the parameter MonitorJMX.Listen. If it reports somthing like 127.0.0.1, jconsole would not be able to connect remotely to MigratoryData Server. To fix this issue, edit /etc/hosts so that hostname resolves to the the address used in MonitorJMX.Listen.
MonitorJMX.Authentication | |
Description | Decide whether to use authentication when connecting to the JMX monitoring service |
Default value | No default value |
Required parameter | Optional |
This parameter can have two values: true or false. If set on true then, in order to access the monitoring via JMX, you will need to use the user name configured with the parameter MonitorUsername and the password configured with the parameter MonitorPassword.
MonitorJMX.Encryption | |
Description | Decide whether to use TLS/SSL encryption when connecting to the JMX monitoring service |
Default value | No default value |
Required parameter | Optional |
This parameter can have two values: true or false. If set on true, then a JMX client will connect to MigratoryData Server through a SSL/TLS encrypted connection. The use of an encrypted connection is especially recommended for the JMX remote monitoring from an insecure network, including Internet.
If MonitorJMX.Encryption is set on true, then the address used in the configuration of the parameter MonitorJMX.Listen must have an certificate entry in the keystore file defined by the parameter KeyStore. The certificate entry in the keystore must be created with the alias jmx as explained in Section 4.5.1.
In this subsection it is assumed that the JMX client used to connect to MigratoryData Server is jconsole. For other JMX clients you can also use the steps below, but you may have to adapt the configuration to suit your specific JMX client.
Supposing the file name of the keystore defined by KeyStore paramater is mykeystore.jks, and supposing the keystore mykeystore.jks includes an certificate entry for the address used in the configuration of the parameter MonitorJMX.Listen having as alias jmx as explained in Section 4.5.1. Then, use the following steps to establish a secure JMX connection:
keytool
-export
-alias jmx
-keystore mykeystore.jks
-rfc
-file temp.cer
keytool
-import
-alias jmx
-file temp.cer
-keystore mytruststore.jks
For the first command above use the password of the keystore as defined by the parameter KeyStorePassword. For the second command above use a new password, say mytruststore-password.
keytool
-genkeypair
-alias jmx
-keyalg RSA
-validity 3600
-keystore clientkeystore.jks
For this command use a new password, say clientkeystore-password.
jconsole
-J-Djavax.net.ssl.keyStore=clientkeystore.jks
-J-Djavax.net.ssl.keyStorePassword=clientkeystore-password
-J-Djavax.net.ssl.trustStore=mystruststore.jks
-J-Djavax.net.ssl.trustStorePassword=mytruststore-password
MonitorHTTP.Listen | |
Description | Address used to listen for HTTP monitoring clients |
Default value | No default value |
Required parameter | Optional |
The format of this addresse is "IP:Port" (e.g. 192.168.1.1:8808) or "Name:Port" (e.g. push.example.com:8808). IPv6 addresses must be enclosed in square brackets
(e.g. [2001:db8::a00:20ff:fea7:ccea]:8808).
Note -- In order to access the HTTP monitoring in MigratoryData Server from a remote machine, please check your firewall settings so that the network traffic from the remote machine is allowed to the address configured by this parameter MonitorHTTP.Listen.
Supposing you have the following monitoring related configuration:
Monitor = HTTP
MonitorUsername = admin
MonitorPassword = pass
MonitorHTTP.Listen = push.example.com:8808
MonitorHTTP.Authentication = true
MonitorHTTP.Encryption = false
Then you can monitor MigratoryData Server by opening the following URL:
http://push.example.com:8808/stats?user=admin&password=pass
Opening the URL above will produce an output with the following format:
<fieldname1>:<value1> <fieldname2>:<value2> ... <fieldnameN>:<valueN>
where each field correspond to one of the following statistics:
applied to one of the following indicators:
for one of the following period of time:
You can also retrieve data in XML and JSON format. Please append a view GET parameter to the URL above with the value xml or json. For example, to retrieve monitoring data in XML format use a URL as follows:
http://push.example.com:8808/stats?user=user&password=pass&view=xml
You can filter the monitoring data by indicator and/or statistic and/or period of time. To do so add to the URL above one or more of the GET parameters listed in Table 4.2.
For example to retrieve the maximum number of concurrent clients in the last 15 minutes, use the following URL written on a single line without line breaks:
http://push.example.com:8808/stats?
user=user&
password=pass&
indicator=ConnectedSessions&
statistic=MAX&
period=Last.15.Minute
Another example. To retrieve the average for all indicators for all periods of time available, use a URL as follows:
http://push.example.com:8808/stats?user=user&password=pass&statistic=AVG
To securely monitor MigratoryData Server over an insecure network such as Internet, configure
MonitorHTTP.Encryption = true
See Section 4.15 for more details. Then use a URL as follows:
https://push.example.com:8808/stats?user=user&password=pass
MonitorHTTP.Authentication | |
Description | Decide whether to use authentication when connecting to the HTTP monitoring service |
Default value | No default value |
Required parameter | Optional |
This parameter can have two values: true or false. If set on true then, in order to access the monitoring via HTTP, you will need to use the user name configured with the parameter MonitorUsername and the password configured with the parameter MonitorPassword.
MonitorHTTP.Encryption | |
Description | Decide whether to use SSL/TLS encryption when connecting to the HTTP monitoring service |
Default value | No default value |
Required parameter | Optional |
This parameter can have two values: true or false. If set on true, then a HTTP monitoring client will connect to MigratoryData Server through a SSL/TLS encrypted connection. The use of an encrypted connection is especially recommended for the HTTP remote monitoring from an insecure network, including Internet.
If MonitorHTTP.Encryption is set on true, then the address used in the configuration of the parameter MonitorHTTP.Listen must have an certificate entry in the keystore file defined by the parameter KeyStore. See Section 4.5.1 to lean how to add a certificate to the keystore.
MonitorPUSH.Stats | |
Description | Decide whether to provide stats logs via a meta-subject |
Default value | false |
Required parameter | Optional |
This parameter can have two values: true or false. If set on true, then a PUSH monitoring client will be able to subscribe to the meta-subjects listed above in Section 4.7 to get the stats logs of the MigratoryData server.
MonitorPUSH.AccessLog | |
Description | Decide whether to provide access logs via a meta-subject |
Default value | false |
Required parameter | Optional |
This parameter can have two values: true or false. If set on true, then a PUSH monitoring client will be able to subscribe to the meta-subject to get the access logs of the MigratoryData server.
MonitorPUSH.MessageLog | |
Description | Decide whether to provide message logs via a meta-subject |
Default value | false |
Required parameter | Optional |
This parameter can have two values: true or false. If set on true, then a PUSH monitoring client will be able to subscribe to the meta-subject to get the message logs of the MigratoryData server.
MonitorPUSH.PublishLog | |
Description | Decide whether to provide publish logs via a meta-subject |
Default value | false |
Required parameter | Optional |
This parameter can have two values: true or false. If set on true, then a PUSH monitoring client will be able to subscribe to the meta-subject to get the publish logs of the MigratoryData server.
LogFolder | |
Description | The folder where the logs will be written |
Default value | logs |
Required parameter | Optional |
If not configured, MigratoryData Server will use the default folder logs relative to the directory path used to start MigratoryData Server.
The log folder can be configured using absolute paths. A value example for this parameter is as follows:
Example of log folder | Platform |
LogFolder = /some/path/mylogs | For Linux/Unix |
LogFolder = C:/some/path/mylogs | For Windows |
LogLevel | |
Description | The level of verbosity of the messages recorded in the logs |
Default value | INFO |
Required parameter | Optional |
The following levels are available:
LogRotateLimit | |
Description | The maximum capacity of a log file expressed in kilobytes (KB), megabytes (MB), or gigabytes (GB) |
Default value | 10 MB |
Required parameter | Optional |
If the log file reaches the capacity provided by this parameter, then MigratoryData Server will automatically create a new log file. The previous log files are preserved on disk up to the number of log files defined by the parameter LogRotateFileCount.
LogRotateTime | |
Description | The time interval at which a new log file will be created expressed in minutes (m), hours (h), days (D), weeks (W), months (M), or years (Y). |
Default value | No default value |
Required parameter | Optional |
For example, in order to record the logs in a separate file every day use:
LogRotateTime = 1 D
To record the logs in separate file every 4 hours use:
LogRotateTime = 4 h
The previous log files are preserved on disk up to the number of log files defined by the parameter LogRotateFileCount.
Note -- This parameter takes precedence over the parameter LogRotateLimit. Therefore, if the parameter LogRotateTime is configured, then the configuration of the parameter LogRotateLimit is ignored.
LogRotateFileCount | |
Description | Limit the number of historical log files created by log rotation |
Default value | 100 |
Required parameter | Optional |
If the number of log files produced by log rotation defined by the parameters LogRotateTime or LogRotateLimit reaches the value of this parameter, then the oldest log file is removed whenever a new log file is created such that the total number of logs files will not exceed the value of this parameter.
DocumentRoot | |
Description | The folder from which MigratoryData Server will serve files |
Default value | html |
Required parameter | Optional |
This parameter is optional, if not supplied the default folder html relative to the directory path used to start MigratoryData Server is used.
Note -- In a production environment, the parameter DocumentRoot should be disabled.
In a production environment, the files of a web application built with MigratoryData APIs should be installed in a standard web server such as Apache. MigratoryData Server has a limited capability to serve web pages, which is only offered to facilitate the development process. The purpose of MigratoryData Server is to push real-time data. Thus, use a web server to provide any static resources necessary for a website such as images, text, and server-side scripting.
PublishSnapshotMessage | |
Description | Specify whether to publish the initial snapshot message or not |
Default value | true |
Required parameter | Optional |
When a client subscribes to a subject, the MigratoryData server will first publish the snapshot message for that subject, and subsequently it will publish any new message for that subject to the client.
See MigratoryData Architecture Guide, chapter Concepts to learn more about the "Snaphot Message" notion.
ClusterPassword | |
Description | Each MigratoryData server in a cluster communicates with the other MigratoryData servers in the cluster, including with itself. This parameter defines the password used by each cluster member for connecting to the other cluster members, including to itself. |
Default value | No default value |
Required parameter | Required |
This password is used only by the MigratoryData servers that form a cluster to connect each other. This password is not used for clients. In order to allow or deny subscriptions and publications for clients, you can enable the entitlement service via the parameter Entitlement.
ClusterDeliveryMode | |
Description | Specify whether you want to use Stanadard Message Delivery or Guaranteed Message Delivery |
Default value | Standard |
Required parameter | Optional |
Define the quality-of-service level for message delivery that your cluster of MigratoryData servers will use. The possible values are: "Standard" (Standard Message Delivery) and "Guaranteed" (Guaranteed Message Delivery)
Using Standard Message Delivery, in the case of a failover reconnection, the client will get the latest (most recent) message available at the moment of the reconnection for each of its subscribed subjects.
Using Guaranteed Message Delivery, in the case of a failover reconnection, the client will get not only the latest (most recent) message for each of its subscribed subjects, but it will also get all the potential messages published during the failover period for each of its subscribed subjects.
See MigratoryData Architecture Guide, chapter Guaranteed Message Delivery, for a complete discussion.
ClusterMemberListen | |
Description | Each MigratoryData server in a cluster communicates with the other MigratoryData servers in the cluster, including with itself. This parameter represents the address used to listen for incoming connections from the other cluster members, including from itself. |
Default value | No default value |
Required parameter | Required |
The format of this listen address is: "IP_Address:Port" (e.g. 192.168.0.1:8801) or "DNS_Name:Port" (e.g. push.example.com:8801) IPv6 addresses must be enclosed in square brackets. For example [2001:db8::a00:20ff:fea7:ccea]:8801.
Note - The port defined by this parameter must be allowed by the firewall for incoming connections from all MigratoryData servers of the cluster.
In addition, the four consecutive port numbers starting with the port number defined by this parameter must be allowed by the firewall for incoming connections from all MigratoryData servers of the cluster.
For example, supposing that this parameter is configured as follows:
ClusterMemberListen = push.example.com:8801
Then the ports 8801, 8802, 8803, 8804, 8805 must be allowed by the firewall for the internal cluster communication between its members.
ClusterMembers | |
Description | Define the cluster by specifying its members. |
Default value | No default value |
Required parameter | Required |
Each cluster member is specified by the listen address defined by its parameter ClusterMemberListen.
For developing and testing purposes, you can deploy a cluster of MigratoryData servers where its members run on the same machine. For example, to define a cluster of three MigratoryData servers all running on the same machine, use
ClusterMembers = 192.168.1.1:7701, 192.168.1.1:8801, 192.168.1.1:9901
For production purposes, you should deploy a cluster of MigratoryData servers where each member runs on a different machine. For example, to define a production cluster of three MigratoryData servers, use something like:
ClusterMembers = 192.168.1.1:8801, 192.168.1.2:8801, 192.168.1.3:8801
Note -- The cluster definition must be identical in all cluster members (the order of the listen addresses must be preserved in the configuration of each cluster member)
ClusterMemberCoordinationFolder | |
Description | A folder used to store internal information useful for cluster coordination |
Default value | coordination_logs |
Required parameter | Optional |
If not configured, MigratoryData Server will use the default folder coordination_logs relative to the directory path used to start the MigratoryData server.
The coordination log folder can be configured using absolute paths. A value example for this parameter is as follows:
Example of transactions folder | Platform |
ClusterMemberCoordinationFolder = /some/path/coordination_logs | For Linux/Unix |
ClusterMemberCoordinationFolder = C:/some/path/coordination_logs | For Windows |
Note -- It is recommended to configure this folder on a local disk instead of a network-attached disk
Entitlement | |
Description | Entitlement type. The possible values are Basic and Custom |
Default value | Basic |
Required parameter | Optional |
Also, if PUSH monitoring is enabled via the parameter Monitor, subscription to the monitoring meta-subjects is allowed only from the clients which authenticate with the token defined by the parameter EntitlementAllowToken.
See MigratoryData Architecture Guide, Chapter Entitlement, to learn more about the Entitlement feature.
EntitlementAllowToken | |
Description | If the Entitlement type is Basic, then only the clients which use the token defined by this parameter are allowed to publish messages |
Default value | No default value |
Required parameter | Required if the Entitlement type is Basic |
If the Entitlement type is Custom, then this parameter is ignored, and the entitlement is done according to your entitlement rules defined by your extension built with MigratoryData Extension API.
RunAsUser | |
Description | Configure MigratoryData Server to run as an unprivileged (normal) user. |
Default value | No default value |
Required parameter | Optional |
Note -- This parameter should be configured only for the Linux platforms.
Supposing "migratorydata" is an existing normal user. Configure MigratoryData Server as follows:
RunAsUser = migratorydata
Now, start the push server as root (this is necessary to be able to bind on the privileged ports 80 or 443). Please note that while running as root, MigratoryData Server will not accept any client connections. Then, MigratoryData Server will drop the root privileges (using the system call setuid) and will automatically switch to the normal user "migratorydata". Only at this time, MigratoryData Server will start to accept client connections.
PublishAllowFromAddressList | |
Description | Define the list of IP addresses allowed for message publication. |
Default value | No default value |
Required parameter | Optional |
If this parameter is configured, then the MigratoryData server will accept message publications only from clients running on any of the IP addresses defined by this parameter.
If this parameter is not set, message publication will be allowed from any client provided however that the client is allowed by the entitlement rules you define (see parameter Entitlement).
Note -- Use only dotted-decimal notation for the IP addresses (no DNS names) to specify the list of IP addresses. For example:
PublishAllowFromAddressList = 192.168.1.100, 192.168.1.101
The advanced parameters of MigratoryData Server are described below.
ClusterCommunication.NIO | |
Description | Specify whether you want to use a NIO thread model for inter-cluster communication |
Default value | false |
Required parameter | Optional |
This parameter is strictly related to the communication between cluster members (and not between clients and server).
By configuring this parameter on true, the MigratoryData server will use non-blocking IO (NIO) operations for inter-cluster communication as well as a simplified thread model (with a configurable number of threads, see the parameter ClusterCommunication.NIO.IoThreads).
Both communications methods (NIO and standard IO) are implemented to provide high performance. Therefore, we recommend to use NIO only if your cluster size is larger than 10 servers.
ClusterCommunication.NIO.IoThreads | |
Description | Define the number of threads to be used for NIO inter-cluster communication |
Default value | 2 |
Required parameter | Optional |
This parameter applies only if the parameter ClusterCommunication.NIO is set on true.
Native.Ssl | |
Description | Use BoringSSL instead of Java SSL implementation for encrypted connections |
Default value | false |
Required parameter | Optional |
The requirements for this option is to use Linux. Use this option to optimize CPU and memory usage.
MaxCachedMessagesPerSubject | |
Description | The number of the most recent received messages from publishers to be stored for each subject |
Default value | 1000 |
Required parameter | Optional |
The number of messages to be stored in memory for each subject.
Note -- This parameter applies only if the Guaranteed Message Delivery feature is enabled (see parameter ClusterDeliveryMode).
SnapshotExpireTime | |
Description | The number of seconds to persist a snapshot message |
Default value | 0 |
Required parameter | Optional |
The MigratoryData server saves a snapshot message in memory for every subject. By default, the snapshot message is persisted in memory for an indefinite period. However, you can specify a limit on the retention time of the snapshot messages using this parameter.
This feature is especially helpful when new subjects are continuously added to the system, but they are not updated after a certain time. Without a time limit, the snapshot messages of these inactive subjects would accumulate in memory, leading to potential memory issues.
Note -- The minimum non-null value of this parameter must be higher than the value of the parameter CacheExpireTime which defaults to 180 seconds.
See MigratoryData Architecture Guide, chapter Concepts to learn more about the "Snaphot Message" notion.
CacheExpireTime | |
Description | The number of seconds a message is held in the cache before it expires |
Default value | 180 |
Required parameter | Optional |
Messages are removed continuously from the cache of each subject, however each message is held in the cache at least the number of seconds defined by this parameter.
Note -- This parameter applies only if the Guaranteed Message Delivery feature is enabled (see parameter ClusterDeliveryMode).
Workgroups | |
Description | The number of groups of clients |
Default value | The number of total CPU cores available |
Required parameter | optional |
In order to better scale on multiprocessor servers the incoming users are separated in groups. This parameter configures the number of groups (every group has a dedicated thread). If not supplied the total CPU cores available is the default value. In most situations it is not recommended to modify the default value.
IoThreads | |
Description | The number of threads used for I/O processing |
Default value | The number of total CPU cores available |
Required parameter | Optional |
If not supplied the number of total CPU cores available is the default value. In most situations it is not recommended to modify the default value.
AccessLog | |
Description | Specify whether to record access logs or not |
Default value | false |
Required parameter | Optional |
The access log records various information about requests to the MigratoryData server, including:
AccessLog.Folder | |
Description | The folder where the access logs will be written |
Default value | logs |
Required parameter | Optional |
If not configured, and the parameter AccessLog is set on true, then MigratoryData Server will use the default folder logs relative to the directory path used to start MigratoryData Server.
The log folder can be configured using absolute paths. A value example for this parameter is as follows:
Example of access log folder | Platform |
AccessLog.Folder = /some/path/mylogs | For Linux/Unix |
AccessLog.Folder = C:/some/path/mylogs | For Windows |
AccessLog.Compression | |
Description | Enable compression of access logs |
Default value | false |
Required parameter | Optional |
In order to reduce the amount of access logs, you can set this parameter on true to enable on-the-fly access log compression.
AccessLog.RotateLimit | |
Description | The maximum capacity of an access log file expressed in kilobytes (KB), megabytes (MB), or gigabytes (GB) |
Default value | 10 MB |
Required parameter | Optional |
If the access log file reaches the capacity provided by this parameter, then MigratoryData Server will automatically create a new access log file. The previous access log files are preserved on disk up to the number of access log files defined by the parameter AccessLog.RotateFileCount.
AccessLog.RotateTime | |
Description | The time interval at which a new access log file will be created expressed in minutes (m), hours (h), days (D), weeks (W), months (M), or years (Y). |
Default value | No default value |
Required parameter | Optional |
For example, in order to record the access logs in a separate file every day use:
AccessLog.RotateTime = 1 D
To record the access logs in separate file every 4 hours use:
AccessLog.RotateTime = 4 h
The previous access log files are preserved on disk up to the number of access log files defined by the parameter AccessLog.RotateFileCount.
Note -- This parameter takes precedence over the parameter AccessLog.RotateLimit. Therefore, if the parameter AccessLog.RotateTime is configured, then the configuration of the parameter AccessLog.RotateLimit is ignored.
AccessLog.RotateFileCount | |
Description | Limit the number of historical access log files created by access log rotation |
Default value | 100 |
Required parameter | Optional |
If the number of access log files produced by access log rotation defined by the parameters AccessLog.RotateLimit or AccessLog.RotateTime reaches the value of this parameter, then the oldest access log file is removed whenever a new access log file is created such that the total number of access logs files will not exceed the value of this parameter.
MessageLog | |
Description | Specify whether to record message logs or not |
Default value | false |
Required parameter | Optional |
The message log records the messages received by the MigratoryData server from client for publication. For each message received, the following information are recorded:
The message log can be used as input for MigratoryData Replayer to "replay" the recorded messages.
MessageLog.Folder | |
Description | The folder where the message logs will be written |
Default value | logs |
Required parameter | Optional |
If not configured, and the parameter MessageLog is set on true, then MigratoryData Server will use the default folder logs relative to the directory path used to start MigratoryData Server.
The log folder can be configured using absolute paths. A value example for this parameter is as follows:
Example of message log folder | Platform |
MessageLog.Folder = /some/path/mylogs | For Linux/Unix |
MessageLog.Folder = C:/some/path/mylogs | For Windows |
MessageLog.Compression | |
Description | Enable compression of message logs |
Default value | false |
Required parameter | Optional |
In order to reduce the amount of message logs, you can set this parameter on true to enable on-the-fly message log compression.
MessageLog.RotateLimit | |
Description | The maximum capacity of an message log file expressed in kilobytes (KB), megabytes (MB), or gigabytes (GB) |
Default value | 10 MB |
Required parameter | Optional |
If the message log file reaches the capacity provided by this parameter, then MigratoryData Server will automatically create a new message log file. The previous message log files are preserved on disk up to the number of message log files defined by the parameter MessageLog.RotateFileCount.
MessageLog.RotateTime | |
Description | The time interval at which a new message log file will be created expressed in minutes (m), hours (h), days (D), weeks (W), months (M), or years (Y). |
Default value | No default value |
Required parameter | Optional |
For example, in order to record the message logs in a separate file every day use:
MessageLog.RotateTime = 1 D
To record the message logs in separate file every 4 hours use:
MessageLog.RotateTime = 4 h
The previous message log files are preserved on disk up to the number of message log files defined by the parameter MessageLog.RotateFileCount.
Note -- This parameter takes precedence over the parameter MessageLog.RotateLimit. Therefore, if the parameter MessageLog.RotateTime is configured, then the configuration of the parameter MessageLog.RotateLimit is ignored.
MessageLog.RotateFileCount | |
Description | Limit the number of historical message log files created by message log rotation |
Default value | 100 |
Required parameter | Optional |
If the number of message log files produced by message log rotation defined by the parameters MessageLog.RotateLimit or MessageLog.RotateTime reaches the value of this parameter, then the oldest message log file is removed whenever a new message log file is created such that the total number of message logs files will not exceed the value of this parameter.
CacheLog | |
Description | Specify whether to record cache logs or not |
Default value | false |
Required parameter | Optional |
If Guaranteed Message Delivery feature is enabled for the MigratoryData cluster, then each cluster member maintains - for each subject - an in-memory cache with the latest messages (for the number of seconds defined by the parameter CacheExpireTime which defaults to 180 seconds, limited however to the number of messages defined by the parameter MaxCachedMessagesPerSubject which defaults to 1000 most recent messages). Moreover, if there is a failure in the cluster, after recovery the cache is automatically synchronized for each subject and cluster member. In this way, for each subject, each cluster member has the most recent messages for that subject.
This cache can be used to retrieve historical messages using the API method subscribeWithHistory. But the main purpose of maintaining a cache is to achieve Guaranteed Message Delivery. For example, if the connection between a client and a cluster member is broken, the client will automatically reconnect to another cluster member. Before resubscribing to its subjects, it will retrieve the messages published during the fail-over time from the cache of the new cluster member.
For every subject, whenever a message is added to the cache of the subject (either during normal operation or cluster recovery), an ADD entry is added to Cache Logs. Whenever a message expires and is removed from the cache, a CLEAN entry is added to Cache Logs. Whenever a client accesses the cache (either via subscribeWithHistory or during fail-over recovery), a GET entry is added to Cache Logs, containing the client identification and the identifiers (i.e. sequence number + epoch number) of the messages provided to that client from the cache.
Cache Logs deal only with client identifiers (IP address + port) and message identifiers (sequence number + epoch number). In order to find an actual client or an actual message, you will have to correlate Cache Logs with Access Logs and Message Logs. So, Cache Logs (correlated with Access Logs and Message Logs) enable to reconstruct the cache of a subject at a given time and how recovery performed for each client. In this way, the Cache Logs will act as a more concrete model where it is much easier to audit the correctness of the system.
CacheLog.Folder | |
Description | The folder where the cache logs will be written |
Default value | logs |
Required parameter | Optional |
If not configured, and the parameter CacheLog is set on true, then MigratoryData Server will use the default folder logs relative to the directory path used to start MigratoryData Server.
The log folder can be configured using absolute paths. A value example for this parameter is as follows:
Example of publish log folder | Platform |
CacheLog.Folder = /some/path/mylogs | For Linux/Unix |
CacheLog.Folder = C:/some/path/mylogs | For Windows |
CacheLog.Compression | |
Description | Enable compression of cache logs |
Default value | false |
Required parameter | Optional |
In order to reduce the amount of cache logs, you can set this parameter on true to enable on-the-fly cache log compression.
CacheLog.RotateLimit | |
Description | The maximum capacity of an cache log file expressed in kilobytes (KB), megabytes (MB), or gigabytes (GB) |
Default value | 10 MB |
Required parameter | Optional |
If the cache log file reaches the capacity provided by this parameter, then MigratoryData Server will automatically create a new cache log file. The previous cache log files are preserved on disk up to the number of cache log files defined by the parameter CacheLog.RotateFileCount.
CacheLog.RotateTime | |
Description | The time interval at which a new cache log file will be created expressed in minutes (m), hours (h), days (D), weeks (W), months (M), or years (Y). |
Default value | No default value |
Required parameter | Optional |
For example, in order to record the cache logs in a separate file every day use:
CacheLog.RotateTime = 1 D
To record the cache logs in separate file every 4 hours use:
CacheLog.RotateTime = 4 h
The previous cache log files are preserved on disk up to the number of cache log files defined by the parameter CacheLog.RotateFileCount.
Note -- This parameter takes precedence over the parameter CacheLog.RotateLimit. Therefore, if the parameter CacheLog.RotateTime is configured, then the configuration of the parameter CacheLog.RotateLimit is ignored.
CacheLog.RotateFileCount | |
Description | Limit the number of historical cache log files created by log rotation |
Default value | 100 |
Required parameter | Optional |
If the number of cache log files produced by log rotation defined by the parameters CacheLog.RotateLimit or CacheLog.RotateTime reaches the value of this parameter, then the oldest cache log file is removed whenever a new cache log file is created such that the total number of cache logs files will not exceed the value of this parameter.
PublishLog | |
Description | Specify whether to record publish logs or not |
Default value | false |
Required parameter | Optional |
If Guaranteed Message Delivery feature is enabled for the MigratoryData cluster, then publish logs record the messages published to and acknowledged by a number of users defined by PublishLog.SampleUsers which are randomly selected from the total number of users.
PublishLog.Folder | |
Description | The folder where the publish logs will be written |
Default value | logs |
Required parameter | Optional |
If not configured, and the parameter PublishLog is set on true, then MigratoryData Server will use the default folder logs relative to the directory path used to start MigratoryData Server.
The log folder can be configured using absolute paths. A value example for this parameter is as follows:
Example of cache log folder | Platform |
PublishLog.Folder = /some/path/mylogs | For Linux/Unix |
PublishLog.Folder = C:/some/path/mylogs | For Windows |
PublishLog.Compression | |
Description | Enable compression of publish logs |
Default value | false |
Required parameter | Optional |
In order to reduce the amount of publish logs, you can set this parameter on true to enable on-the-fly publish log compression.
PublishLog.RotateLimit | |
Description | The maximum capacity of a publish log file expressed in kilobytes (KB), megabytes (MB), or gigabytes (GB) |
Default value | 10 MB |
Required parameter | Optional |
If the publish log file reaches the capacity provided by this parameter, then MigratoryData Server will automatically create a new publish log file. The previous publish log files are preserved on disk up to the number of publish log files defined by the parameter PublishLog.RotateFileCount.
PublishLog.RotateTime | |
Description | The time interval at which a new publish log file will be created expressed in minutes (m), hours (h), days (D), weeks (W), months (M), or years (Y). |
Default value | No default value |
Required parameter | Optional |
For example, in order to record the publish logs in a separate file every day use:
PublishLog.RotateTime = 1 D
To record the publish logs in separate file every 4 hours use:
PublishLog.RotateTime = 4 h
The previous publish log files are preserved on disk up to the number of publish log files defined by the parameter PublishLog.RotateFileCount.
Note -- This parameter takes precedence over the parameter PublishLog.RotateLimit. Therefore, if the parameter PublishLog.RotateTime is configured, then the configuration of the parameter PublishLog.RotateLimit is ignored.
PublishLog.RotateFileCount | |
Description | Limit the number of historical publish log files created by log rotation |
Default value | 100 |
Required parameter | Optional |
If the number of publish log files produced by log rotation defined by the parameters PublishLog.RotateLimit or PublishLog.RotateTime reaches the value of this parameter, then the oldest publish log file is removed whenever a new publish log file is created such that the total number of publish logs files will not exceed the value of this parameter.
PublishLog.SampleUsers | |
Description | Define the number of sample users for publish logs |
Default value | 100 |
Required parameter | Optional |
Define the number of users randomly selected from the total number of users for which to record publish logs and compute latency statistics.
Note -- By configuring this parameter on 0, all users of the system will be used for publish logs. However, recording publish logs for all users and computing latency statistics by asking each user to acknowledge the reception of each message is highly discouraged unless the outgoing throughput of messages published by the server to the users is low enough.
PublishLog.RecordContent | |
Description | Define whether to include the content of the message into the publish logs |
Default value | false |
Required parameter | Optional |
PublishLog.ComputeLatencyStats | |
Description | Compute various statistics about the round-trip latency of messages published to the sample users |
Default value | false |
Required parameter | Optional |
PublishLog.LatencyStatsOutputFrequency | |
Description | Define the frequency to output latency stats |
Default value | 200 |
Required parameter | Optional |
Record in the publish logs various statics about the round-trip latency (i.e. the time necessary for a message to be delivered from the server to a client plus the time necessary to get the reception acknowledgment for the client to the server). The latency stats are computed for all messages published to the sample users, but they are output to the logs only after having processed the number of messages defined by this parameter.
PublishLog.LatencyStatsOnly | |
Description | Record only the latency stats without recording the publish logs |
Default value | false |
Required parameter | Optional |
If this parameter is configured on true, the MigratoryData server will record only the latency stats and will not record the publish logs.
Stats.LogInterval | |
Description | Various statistics are logged at time intervals (seconds) defined by this parameter |
Default value | 60 |
Required parameter | Optional |
The minimum value of this parameter is 5 (seconds).
MaxBatchingSpace | |
Description | The maximum size of the batching in bytes |
Default value | 0 |
Required parameter | Optional |
See MigratoryData Architecture Guide, Chapter Batching, to understand the concept of batching before configuring this parameter. If this parameter is not configured or configured with the default 0 value, then that means batching is disabled.
MaxBatchingTime | |
Description | The maximum time of the batching in milliseconds |
Default value | 0 |
Required parameter | Optional |
See MigratoryData Architecture Guide, Chapter Batching, to understand the concept of batching before configuring this parameter. If this parameter is not configured or configured with the default 0 value, then that means batching is disabled.
CipherListEnabled | |
Description | Enable one or more SSL ciphers in addition to the default JVM ciphers. |
Default value | No default value |
Required parameter | Optional |
The JVM supports a number of ciphers as listed in:
https://docs.oracle.com/javase/8/docs/technotes/guides/security/SunProviders.html
Some of these ciphers are enabled by default. Use this parameter to enable one or more supported ciphers not enabled by default.
For example, to enable the ciphers
configure this parameter as follows:
CipherListEnabled = TLS_DHE_RSA_WITH_AES_128_CBC_SHA, TLS_DHE_DSS_WITH_AES_128_CBC_SHA
CipherListExcluded | |
Description | Exclude one or more SSL ciphers from the default JVM ciphers. |
Default value | No default value |
Required parameter | Optional |
The JVM supports a number of ciphers as listed in:
https://docs.oracle.com/javase/8/docs/technotes/guides/security/SunProviders.html
Some of these ciphers are enabled by default. Use this parameter to disable one or more ciphers enabled by default.
For example, to disable the ciphers
configure this parameter as follows:
CipherListExcluded = TLS_DHE_RSA_WITH_AES_128_CBC_SHA, TLS_DHE_DSS_WITH_AES_128_CBC_SHA
SocketBufferLimit | |
Description | The maximum number of bytes of a client request |
Default value | 65536 (64 KB) |
Required parameter | Optional |
The value of this parameter determines the maximum size (in bytes) of a client requests that MigratoryData Server will accept.
BufferLimit.Send | |
Description | The default initial memory size (in bytes) for outgoing messages |
Default value | 8192 |
Required parameter | Optional |
The default memory size used by MigratoryData Server to create an outgoing message to be sent to the clients is given by the value of this parameter. If the actual size of the message is larger than this default value, then MigratoryData Server will reallocate the necessary memory size.
The default value of BufferLimit.Send is 8192 bytes. You could adjust the value of this parameter based on the common size of your messages to either avoid memory reallocation or optimize memory consumption.
BufferLimit.Receive | |
Description | The default initial memory size (in bytes) for incoming messages |
Default value | 8192 |
Required parameter | Optional |
The default memory size used by MigratoryData Server to create memory space for an incoming message is given by the value of this parameter. If the actual size of the message is larger than this default value, then MigratoryData Server will reallocate the necessary memory size.
The default value of BufferLimit.Receive is 8192 bytes. You could adjust the value of this parameter based on the common size of your messages to either avoid memory reallocation or optimize memory consumption.
Proxy.Type | |
Description | The type of proxy used to connect to the MigratoryData license servers (supported types are SOCKS or HTTP) |
Default value | No default value |
Required parameter | Optional |
If MigratoryData Server is deployed behind a proxy, please configure your proxy type in order to be able to access the MigratoryData licensing servers to validate your evaluation license key. Production license keys do not use remote license verification, thus this parameter should be normally enabled only during the evaluation of MigratoryData Server.
Proxy.Host | |
Description | The address of your proxy used to connect to the MigratoryData license servers |
Default value | No default value |
Required parameter | Optional |
If MigratoryData Server is deployed behind a proxy, please configure the address of your proxy in order to be able to access the MigratoryData licensing servers to validate your evaluation license key. Production license keys do not use remote license verification, thus this parameter should be normally enabled only during the evaluation of MigratoryData Server.
The format of the proxy address is "ip_address:port" (e.g. 192.168.0.1:3128) or "dns_name:port" (e.g. push.example.com:3128).
Proxy.Username | |
Description | The username to authenticate into your proxy to connect to the MigratoryData license servers |
Default value | No default value |
Required parameter | Optional |
If MigratoryData Server is deployed behind a proxy, please configure the user name to authenticate into the proxy in order to be able to access the MigratoryData licensing servers to validate your evaluation license key. Production license keys do not use remote license verification, thus this parameter should be normally enabled only during the evaluation of MigratoryData Server.
Proxy.Password | |
Description | The password to authenticate into your proxy to connect to the MigratoryData license servers |
Default value | No default value |
Required parameter | Optional |
If MigratoryData Server is deployed behind a proxy, please configure the password to authenticate into the proxy in order to be able to access the MigratoryData licensing servers to validate your evaluation license key. Production license keys do not use remote license verification, thus this parameter should be normally enabled only during the evaluation of MigratoryData Server.
Extension.Presence | |
Description | Specify whether to enable presence extension or not |
Default value | false |
Required parameter | Optional |
To enable a presence extension built with MigratoryData Presence API and deployed under the folder extensions of your MigratoryData installation. See Developer's Guide and Reference Manual of MigratoryData Presence API to learn how to build a presence extension.
Extension.Presence.Subject | |
Description | Subject used to replicate presence updates across the cluster |
Default value | /__migratorydata__/presence |
Required parameter | Optional |
Whenever a user connects to or disconnects from a cluster member, a presence update is provided to the presence extension of that cluster member. Moreover, this presence update is replicated across the cluster such that the presence update to be provided to the presence extension of each cluster member.
Internally, MigratoryData uses its client library to replicate presence updates across the cluster. Each cluster member subscribes to the subject defined by this parameter. Whenever a user connects to or disconnects from a cluster member, that cluster member publishes a presence update on the subject defined by this parameter.
Extension.Presence.EntitlementToken | |
Description | Entitlement token used to replicate presence updates across the cluster |
Default value | the value of the parameter EntitlementAllowToken |
Required parameter | Optional |
As detailed in Section 5.51 above, MigratoryData uses its standard messaging to replicate presence updates across the cluster.
The default value of this parameter is the value of the parameter EntitlementAllowToken. In this way, if the value of the parameter Entitlement is Basic, then there is no need to configure this parameter, simply use its default value. Otherwise, if the value of the parameter Entitlement is Custom, then your custom entitlement rules should entitle subscriptions and publications on the subject defined by the parameter Extension.Presence.Subject for the clients using the entitlement token defined by this parameter.
MaxBandwidthRate | |
Description | Specifies the maximum bandwidth rate - in megabytes per second (MB/s) - that MigratoryData server is allowed to consume from the available network bandwidth capacity to push messages to clients |
Default value | 0 |
Required parameter | Optional |
This parameter specifies the maximum bandwidth rate - in megabytes per second (MB/s) - that MigratoryData server is allowed to consume from the available network bandwidth capacity to push messages to clients. If zero is specified, the bandwidth rate will not be limited (this is the default value).
This parameter is particularly useful for use cases where, occasionally, the outgoing message throughput is higher than the available network bandwidth capacity. Supposing the MigratoryData server is installed on a machine with a network bandwidth capacity of a 1 Gbps (i.e. 125 MB/s), and supposing there are one or more large messages to be delivered to a large number of users, producing an outgoing data volume of 1250 MB. By configuring MaxBandwidthRate at 125 MB/s, the assumed available network capacity, the MigratoryData server will be able to deliver that volume of data during 10 seconds, and then operate normally, with low message latency. Otherwise, without configuring MaxBandwidthRate, the system might be severely impacted - as the capacity of the kernel's socket buffers could be exhausted.
Note that configuring MaxBandwidthRate could be useful not only for the use cases discussed above. It can be used as a protection against the deliberate or accidental exceeding of the available network bandwidth capacity for any other use case.
For most cases, MaxBandwidthRate should be configured to be equal to the available network bandwidth capacity.
Note -- With a LAN-based deployment, MaxBandwidthRate can be configured at 125 (MB/s) on a 1 GbE network, and respectively at 1250 (MB/s) on a 10 GbE network. However, with an Internet-based deployment, please check with your network provider to determine the bandwidth capacity available for your deployment.