J06.09 Release Version Update Compendium

2 Operating System
OSS Network File System (NFS) and ACLs
Beginning with the J06.09 RVU, you can use OSS Network File System (NFS) Version 2 clients
to access OSS objects protected by ACLs that contain optional ACL entries if the NFSPERMMAP
attribute value for the fileset that contains the object is set to a value other than DISABLED.
The NFSPERMMAP attribute value selects the algorithm used to map the OSS ACL permissions
for the object to the standard permissions (rwxrwxrwx) expected for the object by NFS V2 clients.
Write permissions are always enforced on the NonStop server using the actual standard OSS
permissions or OSS ACL permissions (if present) on the object.
The NFSPERMMAP attribute is a new attribute of the OSS fileset and is set using Subsystem
Control Facility (SCF) commands. The default setting for the NFSPERMMAP fileset attribute is
DISABLED. To allow NFS V2 clients to access objects in the fileset that are protected by optional
OSS ACL entries, you must change the NFSPERMMAP fileset attribute to RESTRICTIVE,
PERMISSIVE, or UNMODIFIED.
For more information about this feature, see the Open System Services Management and Operations
Guide, the Open System Services Management and Operations Guide, and the acl(5) reference page
either online or in the Open System Services System Calls Reference Manual.
Migration Considerations
The default setting for the NFSPERMMAP fileset attribute is DISABLED. All attempts by NFS
V2 clients to access objects protected by optional OSS ACL entries are denied. This behavior
matches the behavior for earlier RVUs.
To allow NFS V2 clients to access objects in the fileset that are protected by optional OSS ACL
entries, you must change the NFSPERMMAP fileset attribute to RESTRICTIVE, PERMISSIVE,
or UNMODIFIED.
Fallback Considerations
The setting of the NFSPERMMAP fileset attribute is ignored; the behavior is as if it had been set
to DISABLED.
New NonStop Operating System TLE Limit
Beginning with RVU J06.09, the NonStop operating system introduces a new per-process time
list element (TLE) limit. This limit can be set by a super group user by issuing the following SCF
command:
ALTER SUBSYS $ZZKRN, TLE_LIMIT <n>
where n is 101 to 20000. If no limit is set, the limit defaults to 20000. No application process is
allowed to allocate TLEs beyond this limit. An EMS event is generated when a process tries to
cross this limit. The format of the EMS message is:
“TANDEM.NSK.H00 000129 TLE Limit 2000 reached for PIN 225 Guardian filename \NSK01.$SYSTEM.TEST.OTES”
The limit can be altered online using the above-mentioned SCF command. The current limit can
be queried by issuing the SCF INFO SUBSYS $ZZKRN command or by programmatically calling
PROCESSOR_GETINFOLIST_ with an attribute value 79. For more information, see the SCF
Reference Manual for the Kernel Subsystem.
Migration Considerations
This feature is always enabled if RVU J06.09 is installed.
OSS Network File System (NFS) and ACLs 11