This section describes different specifications between PRIMECLUSTER in the non-global zone and PRIMECLUSTER in the global zone.
When migrating PRIMECLUSTER (installing a new PRIMECLUSTER) from the old Solaris 8 or Solaris 9 environment where PRIMECLUSTER was not used to new OSLC environment.
However, this section does not cover the contents described from "13.1.1 Range of Support" to 13.1.3 Resource Configuration."
Note
When migrating PRIMECLUSTER from the 4.2A00 or earlier version to OSLC environment, see the manual of the PRIMECLUSTER version used in the migration source.
The files that define the mount point of the file system used by the Fsystem resource and the entry of NFS are different between in the non-global zone and in the global zone.
Replace /etc/vfstab.pcl file and /etc/dfs/dfstab.pcl file, which are described in this manual, with etc/vfstab file and /etc/dfs/dfstab file instead for PRIMECLUSTER in the non-global zone.
The default value of the environment variable RELIANT_SHUT_MIN_WAIT, which specifies the timeout interval of the hvshut command, is different between in the non-global zone and in the global zone.
The default value of this environment variable RELIANT_SHUT_MIN_WAIT, which specifies the timeout interval of the hvshut command, is 900 (seconds) instead of INTMAX (seconds) for PRIMECLUSTER in the non-global zone. The command times out in 900 (seconds) if this environment variable is unchanged from the default value.
The default value of the RMS local environment variable HV_CONNECT_TIMEOUT is different between in the non-global zone and in the global zone.
The default value of HV_CONNECT_TIMEOUT is 5 (seconds) instead of 30 (seconds) for PRIMECLUSTER in the non-global zone.
The RMS message (SYS, 8) that is logged in the syslog file is different between in the non-global zone and in the global zone.
The RMS message (SYS, 8) that is logged in the syslog file is as follows.
PRIMECLUSTER in the global zone:
(SYS, 8): ERROR: RMS failed to shut down the host <host> via a Shutdown Facility, no further kill functionality is available.
The cluster is now hung. An operator intervention is required.
PRIMECLUSTER in the non-global zone:
(SYS, 8): ERROR: RMS failed to shut down the host <host> via a Shutdown Facility, no further kill functionality is available. The cluster is now hung.
Severity of the following RMS wizard message is different between in the non-global zone and in the global zone.
"cannot grab mount lock for dostat() check_getbdev(), returning previous state"
The message "cannot grab mount lock for dostat() check_getbdev(), returning previous state" is displayed as WARNING instead of NOTICE for PRIMECLUSTER in the non-global zone.
The port number used by RMS is changed.
The port number 11111 is used for PRIMECLUSTER in the global zone while the number is not used for PRIMECLUSTER in the non-global zone.
