i installed a 3-node cluster on my laptop (ultra stable slackware 10 host, 3 customized slackware virtual machines) with openmosix kernel 2.4.26.

then i installed sqlanywhere 9 and ran two databases on the first node.

since it is just a virtual machine, the load on the first node easily reaches 5.0+, but the dbsrv processes never migrate.

there is tremendous traffic between the three nodes (localized to one subnet, of course), and some other processes (not dbsrv9) do migrate.

i try to force the dbsrv9 processes to migrate using the mosctl commands "mosctl runon `ps -ef | grep dbsrv | awk '{ print $2 }'" and "echo 1 > /proc/pid/lock" but they do not.

is there a way to force dbsrv process to migrate or is this inherent in the programming design of the software. as an added question, does dbsrv10 support clustering? any plans for dbsrv11?