Flag

Definition

SpaceName=

This flag determines the specific Space that the Magic xpi server should connect to.

LookupGroupName=

This flag is used by the Magic xpi server to connect to a specific group within the Space that was identified by the SpaceName= flag (above).

The value for this flag should be identical to the LOOKUPGROUPS value defined in the setenv.bat file (located in the GigaSpaces-xpi\bin folder). If the Space is loaded using a LOOKUPGROUPS value that is different from the one defined by the LookUpGroupName= flag, the project will not run in the Space.

LookupLocators=

This flag holds the details of the locators that you defined in the GigaSpaces Configuration screen's Use the following locators (Unicast) parameter. These details are shown as a comma-delimited <host:port> list of the locators' services.

StartProjectFileName=

This flag defines the name of the startup XML for the Magic xpi Studio, Monitor, and Debugger. When a project is started, the Monitor and the Debugger search for this XML under the specific project folder.

The default value for this flag (and if this flag does not exist) is start.xml.

Flag

Definition

TriggersBufferSize=

To prevent asynchronous triggers from flooding the Space with unhandled flow request messages, each trigger maintains a buffer queue (default of 10 messages) that will be used to prevent the creation of new events if the queue exceeds the predefined size.

To overwrite the default value, add this flag to the [MAGICXPI_GS] section of the Magic.ini file, and then set the new trigger default buffer value as required.

The trigger will stop polling once it finds that the amount of unhandled messages in the Space is equal to or greater than the buffer value.

WorkerReadDeeperMessagesfirst=

This flag is used to tell Magic xpi to prioritize messages with a higher depth when reading messages from the Space. This ensures that a business process (a process starting from a root message that may have parallels in various levels) is completed before a new one starts.

The default value for this flag (and if this flag does not exist) is Y.

DIR_SCANNER_ORDERED=

This flag is used to control the message ordering of Directory Scanner triggers.

When this flag is set to Y, the trigger behaves synchronously and will not select the next file until the previous one was processed or aborted.

In addition, recovery must be set on the flow to make sure that even if there is a crash, the flow will end in a controlled way.

CheckServerEntryInspaceRetrytimes=

The Magic xpi engine loads the ServerData object from the Magic Space. In case it fails to load, it will retry for the defined number of times before terminating the engine. The retry count by default is set to 10.

To overwrite the default value, add this flag to the [MAGICXPI_GS] section of the Magic.ini file, and then set the new retry count as required. If the retry count is less than or equal to zero it will be defaulted to 10.

Each retry will be logged at the Debug level along with the retry number in the log file.

IBM_I_ORDERED=

This flag is used to restrict the IBM I trigger to handle one message at a time; thereby conforming to the order (FIFO/FILO) set to the Data Queue. To take effect, add this flag to the [MAGICXPI_GS] section of the Magic.ini or ifs.ini file with value set to Y.

When this flag is set to Y, the trigger behaves synchronously and receives the messages as per the sequence defined on the IBM i component trigger configuration.

PauseOnCriticalConditions

This flag is used by the Magic xpi server for handling the critical state of the system.

The critical conditions are known to lead to data inconsistency or duplicate processing of business event/trigger. On occurrence of any of the critical condition the system will pause execution of new requests thereby reducing the risk of inconsistency and duplication issues.

For more information on critical alerts, click here.

If the flag is set to Y, the xpi server will not accept any new flow requests when the system reaches a critical state like Redo logs are exceeded, Split-Brain condition or database is inaccessible. The server will pause and wait until the critical state is resolved and the system is healed. The threads which are currently processing will continue to run and will not be paused.

The server pause time will be executed based upon an increasing delay algorithm at an interval of 30, 60, with maximum value of 120 minutes. The events will be logged and the administrator will be communicated with an email before every server pause. When the situation is resolved and back to normal, the administrator will be sent an email about the restoration of normal behavior.

The default value is N. The user has to set the value of the flag to Y, to enable this behavior.

Any event of the processing being paused and resumed will be logged into the Gigaspaces logs. The log levels are controlled using the logger "com.magicsoftware.xpi.server.spacebridge”. Refer the Logback Configuration page, for further information.

(Since version: 4.13)