SnappyData CLASSPATH - padogrid/padogrid GitHub Wiki

◀️ KryoSerializer Generator 🔗 SnappyData Docker Compose ▶️

The SnappyData class path is defined in the following order:

$CLASSPATH
$PADOGRID_WORKSPACE/clusters/<cluster>/plugins/*
$PADOGRID_WORKSPACE/clusters/<cluster>/lib/*
$PADOGRID_WORKSPACE/plugins/*
$PADOGRID_WORKSPACE/lib/*
$PADOGRID_HOME/snappydata/plugins/*
$PADOGRID_HOME/snappydata/lib/*
$PADOGRID_HOME/lib/*
$SNAPPYDATA_HOME/jars/*

To see member startup details, run the start_member -simulate command, which does not run the member but simulates the startup sequence that includes displaying the order of CLASSPATH, Java options, and arguments.

start_member -simulate

Note that CLASSPATH and JAVA_OPTS can be set in setenv.sh at each directory level to include additional paths or override any paths set in the subsequent directory levels. The setenv.sh files are located as follows:

$PADOGRID_WORKSPACE/clusters/$CLUSTER/bin_sh/setenv.sh
$PADOGRID_WORKSPACE/setenv.sh
$PADOGRID_HOME/bin_sh/setenv.sh

SnappyData has three distinct types of JVM instances: locators, leaders, and servers. In PadoGrid, servers are referred as members. You can configure each type in the cluster's setenv.sh file.

cd_cluster
vi bin_sh/setenv.sh

In the setenv.sh file you will find three environment variables as follows. Set them in the standard SnappyData argument list format, i.e., "-opt1=val1 -opt2=val2 ...".

LOCATOR_OPTS=""
LEAD_OPTS=""
MEMBER_OPTS=""

◀️ KryoSerializer Generator 🔗 SnappyData Docker Compose ▶️